Cannabis Keyword List: US and Canada, Real Volumes

Most cannabis keyword lists you'll find are the same forty terms rearranged. Somebody typed "dispensary near me" into a free tool, screenshotted the top ten, wrapped it in 800 words, and published it. That list is useless for planning a site, because a dispensary in Tulsa and a delta-8 brand shipping nationally and an MSO with 90 storefronts need three completely different keyword universes, and none of them are covered by forty terms.

This one is different in three ways. Every keyword below carries a measured monthly search volume from the Semrush US database, pulled on September 7, 2026. Every category is split by search intent, so you know whether a term belongs on a store page, a blog post, or nowhere at all. And the list includes the categories other lists skip entirely: brand and competitor terms, licensing and B2B, accessories, cultivation, and the legality queries that drive more cannabis traffic than any product term.

What's in the list

Sixteen categories, 1,041 measured keywords, organized the way a cannabis site actually gets built:

Skip to whichever one matches what you sell. If you run a single storefront, categories 2, 3 and 10 are where your revenue lives and the rest is content marketing. If you're a national CBD or hemp-derived brand, categories 8 and 12 are the business and category 2 is a distraction.

How the numbers were produced, and what they're worth

The volumes come from Semrush's US database via broad-match (phrase_fullsearch), batch lookup (phrase_these) and question reports (phrase_questions), run September 7, 2026. Difficulty is Semrush's keyword difficulty index. Intent is Semrush's classification: informational, navigational, commercial, transactional.

Three honest caveats before you build a budget on any of this.

Search volumes from any vendor are modeled estimates, not counts. Semrush clusters close variants and rounds hard, which is why so many terms in the tables below land on 12,100, 14,800, 18,100 and 22,200. Those are bucket boundaries, not measurements. Treat every figure as an order of magnitude and a ranking signal against the other terms in the same table, not as a traffic forecast. If a keyword shows 40,500 and another shows 4,400, the first is roughly ten times bigger. That's the claim the data supports.

Keyword difficulty is a vendor model too. It's directionally useful for sorting a list of 1,200 terms into "obvious first" and "obvious later," and it should never be the reason you skip a keyword that matches what you sell.

Some of these numbers are contaminated, and I've flagged the worst offenders inline. That contamination is worth a section of its own, because it's the single biggest reason cannabis keyword research goes wrong.

The homonym problem nobody warns you about

Pull "weed" as a broad match in the US database and the top result is how to get rid of weeds at 246,000 a month. Below it: weed eater at 74,000, weed wacker at 74,000, stihl weed eater at 60,500, spruce weed killer at 49,500, horny goat weed at 60,500. Lawn care and herbal supplements outrank the actual plant on its own seed term.

It gets worse. "Edibles" returns edible arrangements at 550,000 and how to make edible cookie dough at 110,000 before it returns anything cannabis. "Strain" returns muscle strain at 33,100, eye strain, repetitive strain injury, and new covid strain. "Bong" returns Bong Joon Ho at 49,500 and bongo drums. "Joint" at 40,500 is mostly anatomy and construction. "Grinder" at 135,000 is mostly angle grinders and a dating app.

A cannabis keyword list that reports those raw numbers is lying to you, and quite a few published ones do. The fix is mechanical: never trust a single-word cannabis seed, always validate with a modifier (weed dispensary, cannabis edibles, weed strains, water bong, weed grinder), and check the SERP before you commit a page to a term. Every number in the tables below has been filtered for this, and where a term is still partly contaminated I say so in the row notes.

1. Head terms and category anchors

These are the terms that define the vertical. Almost none of them are winnable for a normal operator, and almost all of them are worth understanding, because they tell you what the category's vocabulary actually is and which of your pages needs to carry which noun.

The important read here is that cannabis, weed and marijuana are not interchangeable in search. They pull different volumes, different difficulty, and different intent. Cannabis at 201,000 skews clinical and industry. Weed skews consumer and casual. Marijuana at 60,500 skews legal, medical and news. A page that uses only one of the three is invisible to two thirds of the category, which is the core argument behind contextual density and keyword variation coverage.

2. Near-me and local intent

This is where dispensary revenue comes from, and it's the category where most operators still lose to Weedmaps, Leafly and Dutchie-hosted menus rather than to other dispensaries. Recovering those queries from the marketplaces is the first thing dispensary SEO has to fix. Every term here is transactional. Someone typing dispensary open near me is standing up, keys in hand.

Two structural notes before the table. First, the "near me" family does not need a page per phrasing. Google resolves dispensary near me, dispensaries near me, weed dispensary near me and cannabis dispensary near me to substantially overlapping local results, so one well-built store page plus a Google Business Profile carries the whole family. What you actually optimize for is the entity behind the query, not each string.

Second, the Spanish-language variants are real and almost nobody targets them. Dispensary cerca de mi at 18,100 and smoke shops cerca de mi at 22,200 are sitting there in the US database with no serious competition. If you operate in Arizona, Nevada, Illinois, California, Texas or Florida, that's a page.

City-level dispensary terms

The near-me family has a city-modified twin for every metro you operate in. These are the measured US volumes for twenty of them, and the pattern (dispensary + city) extends to any market. Canadian city terms are thin in the US database and should be re-pulled in the ca database before you plan against them.

Toronto at 70 and Calgary at 50 in the US database are not real readings of Canadian demand. They're what Americans search. Anyone planning Canadian city pages should pull the ca database, and anyone planning Ontario coverage should look at how the Toronto cannabis SEO page is structured before building more.

State-level dispensary terms

State-modified retail terms sit between the national head terms and the city terms, and they're the natural target for a state hub page that links down to individual store locations. Michigan at 22,200 is the largest of them by a wide margin.

3. Dispensary and retail store terms

Retail terms without the "near me" attached. These behave differently: they're a mix of navigational (someone looking for a specific chain), commercial (someone comparing), and informational (someone who doesn't know what a dispensary is). The deals and hours cluster is the most underserved part of this table, and it's the cheapest traffic in cannabis retail. Dispensary deals near me at 14,800 with a difficulty of 29 is a weekly-updated page, not an article.

4. Strain keywords

Strain pages are the highest-volume, lowest-difficulty content in cannabis, and they're the reason Leafly and AllBud dominate the vertical. Look at the difficulty column: ice cream cake strain at 18,100 with a KD of 12. Apple fritter strain at 14,800 with a KD of 10. Pink runtz strain at 12,100 with a KD of 7. These are numbers you don't see anywhere else in a mature vertical.

The catch is that a strain page is only worth building if you can say something the strain databases can't. A dispensary that stocks the strain, prices it, photographs the actual jar, and reports its own lab COA has a page nobody else can copy. A dispensary that paraphrases the Leafly entry has a duplicate. That distinction is the entire game in this category.

Watch the contamination here too. Bare strain at 27,100 is mostly muscle strain, eye strain and covid strain. Always keep the modifier.

5. Flower, pre-rolls and concentrates

Product-format terms. These map to category pages on a menu, not to blog posts, with one exception: the "what is" and "vs" queries inside this group are informational and belong in a buying guide that links to the category page.

Pre-rolls are the standout. Pre rolls at 14,800 and infused pre roll at 5,400 come with a long tail of for sale variants at low difficulty, which is unusual for a transactional term. The concentrates group has the same shape: live resin diamonds, live resin badder for sale, sauce live resin for sale all sit in the single-digit to low-teens difficulty range.

6. Vapes, carts and pens

The vape category is where the vocabulary splits hardest between what the industry says and what customers type. Nobody searches "vaporizer cartridge." They search weed carts, thc pen, dab pen, 510 battery. If your product pages use industry nouns and your customers use street nouns, you don't rank, and no amount of link building fixes a vocabulary mismatch.

Note the battery and troubleshooting cluster at the bottom. Battery for weed cart, do weed carts expire, how to open a weed cart, weed cart disposable are support queries with almost no competition, and they bring in exactly the person who is about to buy a replacement.

7. Edibles, gummies and drinks

Two things about this category. First, thc drinks at 74,000 is the fastest-growing product term in the entire list and it barely existed three years ago. The beverage cluster (thc seltzer 14,800, thc infused drinks 9,900, thc beverages 9,900, cannabis infused drinks 5,400, best thc drinks 6,600) is a category page most operators haven't built yet. If you sell hemp-derived beverages, that's the opening.

Second, bare edibles at 60,500 is heavily contaminated by Edible Arrangements and edible cookie dough, and gummies at 33,100 is mostly vitamins. Use thc gummies, weed gummies, cannabis edibles, edibles weed. The dosage and duration questions (how long do edibles last, how long do edibles last in your system) are the highest-intent informational terms in cannabis retail, because the person asking has already bought.

8. CBD, hemp and alternative cannabinoids

This category is a separate business from dispensary retail. It ships nationally, it lives or dies on ecommerce SEO rather than local pack visibility, and its keyword universe is driven by regulatory arbitrage: delta-8, THCA and HHC exist as search categories because of a gap in the 2018 Farm Bill, and the volumes move whenever a state moves.

That volatility is the thing to plan around. Texas thc ban at 12,100 and hemp ban at 9,900 are not evergreen queries, they're event-driven spikes that will look completely different in six months. Build the ban and legality pages knowing they need maintenance, and don't build a whole business on a cannabinoid whose legal status is one bill away from changing.

Also watch cbd in broad match: osaka cbd, christchurch cbd new zealand, weather melbourne cbd and cbd book shops are central business district and bookstore queries, not cannabidiol.

9. Medical cannabis and card keywords

The medical card cluster is state-fragmented and that fragmentation is the opportunity. There is no national winner for how to get a medical marijuanas card because the answer is different in every state, which means a clinic or telehealth operator with one strong state page can outrank a national aggregator inside its own market.

Note the spelling. Search demand overwhelmingly uses "marijuanas" with the incorrect plural: medical marijuanas card variants carry more volume than the grammatically correct forms. That's the vocabulary the market uses, and a page that refuses to use it forfeits the traffic. Write both.

10. Legality, laws and state-level queries

Add up the is weed legal in [state] family and it clears 250,000 searches a month in the US alone. That's more than most product categories on this list, and dispensaries almost never build for it because it doesn't look like a money keyword.

It is one. Somebody asking whether weed is legal in Tennessee is either about to buy or about to drive to a state where they can. A legality page that answers the question honestly, gets updated when the law moves, and links to your store or your delivery zone converts at a rate that surprises operators the first time they measure it. The same page also earns the links and the citations that carry the rest of the domain, which matters in a vertical where paid acquisition is closed off.

One caution on the volumes here: legality search spikes hard around legislative events, and Semrush reports a trailing average. Marijuana rescheduling at 22,200 and cannabis rescheduling at 8,100 reflect a specific news cycle. Treat those two as unstable.

11. Question and informational keywords

Blog territory. Every term here is a page, and none of them belong on a product page. The value of this category isn't direct revenue, it's that these queries are how a cannabis site earns topical coverage in a vertical where you can't buy your way in.

Group them by cluster rather than by volume when you build. The detection cluster (how long does weed stay in your system and its dozen phrasings), the dosage cluster (how much is an eighth of weed, how many grams in a quarter), the effects cluster (why does weed make you hungry, why does weed make you paranoid), the comparison cluster (indica vs sativa and friends), and the preparation cluster (how to roll a joint, how to make cannabis oil). Each cluster is one strong page with the variants placed inside it, not ten thin pages.

12. Brand and competitor keywords

The category every other cannabis keyword list leaves out, and the one with the most volume in it. Trulieve and Curaleaf each pull 301,000 a month. Rise dispensary pulls 246,000. Puffco pulls 201,000. Weedmaps pulls 301,000. Brand demand in cannabis is larger than category demand, because a customer who has found a product they like searches for that product by name from then on.

There are three ways to use this table, and only two of them are a good idea.

Own your own brand terms. If you're an MSO or a multi-store operator, your brand plus location (trulieve near me at 33,100, curaleaf near me at 27,100, rise dispensary near me at 18,100) is the highest-converting query set you have, and it's frequently lost to Weedmaps, Leafly and Dutchie menu pages that outrank the operator's own store page. Fixing that is usually the fastest win available in cannabis SEO.

Carry competitor product names as stocked inventory. If you actually sell Stiiizy, Wyld, Wana, Jeeter, Raw Garden or Kiva, a menu page naming the brand is legitimate, accurate, and rankable. This is not conquesting, it's inventory.

Bidding on or targeting a competitor's brand with content you can't back up is the third option and it isn't worth it in a vertical this closely watched. Cannabis advertising rules vary by state and several prohibit claims that imply an association you don't have. Check your state's rules before you build a comparison page around a licensed competitor's name.

13. Cultivation and home-grow keywords

Small volumes, tiny difficulty, and a buyer at the end of most of them. Home-grow content is worth building if you sell seeds, nutrients, tents, lights or equipment, and it's worth building even if you don't, because it earns the informational coverage that dispensary menus can't.

Cannabis seeds at 14,800 and marijuana seeds at 14,800 are the commercial anchors. Everything below them is a how-to page.

14. Accessories, smoke shop and head shop

Accessories are the part of the vertical where a normal ecommerce site can still beat a licensed operator, because nothing here is a controlled substance and the advertising restrictions are looser. The volumes are large and the difficulty is low: nectar collector at 22,200 with a KD of 15, herb grinder at 12,100 with a KD of 26, smoke buddy at 12,100 with a KD of 19, hemp wraps at 8,100 with a KD of 22.

Contamination warning on three terms. Grinder at 135,000 is mostly power tools and a dating app. Lighter at 60,500 is mostly generic. Hookah at 135,000 is a different product category with its own audience, worth a page only if you actually sell it.

15. B2B, licensing and cannabis industry services

The buyers here are operators, not consumers, and the CPCs tell you exactly how much those buyers are worth. Dispensary pos at 170 searches a month carries a $19.81 CPC. Dispensary software at 210 carries $18.87. Cannabis attorney at 480 carries $16.73. Seed to sale software at 210 carries $15.36. 280e at 590 carries $14.88.

Those are the highest commercial-value keywords in the entire cannabis vertical, and they have two-digit search volumes. This is the shape of every B2B cannabis category: almost no traffic, enormous deal size. If you sell to operators, you are not building for volume, you are building maybe forty pages that each get thirty visits a month from someone with a licence and a budget.

The same logic applies to agency terms. Cannabis seo at 1,900, dispensary seo at 720 with an $11.89 CPC, cbd seo at 1,900 with a $5.85 CPC. Small numbers, and a single dispensary group that converts pays for the whole cluster.

16. Culture, pricing, deals and online ordering

The last category is the one that turns browsers into customers, and it's almost entirely uncontested. Pricing questions carry real volume at low difficulty: weed measurements at 8,100 with a KD of 52, zip of weed at 6,600 with a KD of 17, dime bag at 6,600 with a KD of 27, quarter of weed at 2,400. A dispensary that publishes a plain pricing and measurements page owns a cluster nobody is fighting for.

The deals and loyalty terms at the bottom have two-digit volumes, which is the point. They aren't traffic, they're the pages your existing customers land on when they search your brand plus "deals," and they're trivially cheap to build.

Munchies at 60,500 and 420 meaning at 22,200 are culture terms with mass reach and no commercial intent. They're link bait, and they should be treated as such rather than dressed up as conversion pages.

Turning a thousand keywords into a site

A list this size is a liability if you treat it as a to-do list. A thousand thin pages is how cannabis sites end up cannibalizing themselves, with four URLs competing for dispensary near me and none of them winning.

The mapping that works is roughly this. One store page per physical location, carrying the near-me family, the city term, hours, menu and deals, with the location marked up in LocalBusiness schema. One category page per product format you actually stock, carrying the format vocabulary in both industry and street registers. One page per strain you carry, and only strains you carry, with your own photography, your own price, and your own COA. One legality page per state you operate in or deliver to, maintained on a calendar rather than written once. One cluster page per informational group, absorbing the variant phrasings rather than spawning a page for each. And, if you sell to operators, forty B2B pages that will never bring volume and will bring buyers.

What decides whether any of those pages ranks is coverage of the category's vocabulary, not repetition of the head term. A dispensary page that says "dispensary" sixty times has told Google one thing sixty times. A page that carries dispensary, cannabis store, weed shop, recreational, medical, budtender, menu, delivery, pre-roll, flower, edibles, carts and the strain names on the shelf has told Google fifteen things, and it can rank for all fifteen. That's the measurable part, and it's what our contextual density method is built to score.

Why cannabis makes organic search non-optional

Google Ads doesn't accept cannabis. Meta doesn't accept cannabis. TikTok doesn't accept cannabis. Programmatic is a patchwork of state rules and publisher policies that change without notice. What's left is organic search, local pack visibility, email, and the marketplaces that sit between you and your own customers and charge you for the privilege.

That's the whole argument for taking a list like this seriously. In an unrestricted vertical a bad keyword strategy costs you efficiency, because you can buy your way past it. In cannabis a bad keyword strategy costs you the customer, because there is nothing to buy your way past it with. It's also why the marketplace terms in category 12 matter so much: every dispensary that lets Weedmaps outrank its own store page is renting access to traffic it could own.

Where this list will be wrong in six months

Three parts of it are unstable and you should plan to re-pull them. The hemp-derived cannabinoid group in category 8 depends on federal and state legislation that is actively moving, and terms like texas thc ban and hemp ban are event spikes rather than baselines. The rescheduling cluster in category 10 is news-driven. And the strain names in category 4 turn over faster than anything else in the vertical, because a strain that dominates search in 2026 may not be on a shelf in 2027. Blue Dream has held for a decade. Super Boof probably won't.

The rest, meaning the near-me family, the retail vocabulary, the product formats, the legality questions and the B2B terms, is stable enough to build a site architecture on.

What changes in Canada

Canada legalised nationally in 2018 and sells through provincial systems, so the Canadian list is shaped by geography rather than by legality. The US map carries a state legality cluster, a medical card cluster and a state hub cluster; all three are gone from the Canadian map, which runs 40 clusters against the American 44. What is left is retail, product and education: 541 terms clearing 50 searches a month, measured on Semrush CA in September 2026.

The volumes are a fraction of the American ones and the ratio is not even. Dispensary near me runs 90,500 a month in Canada against 1,830,000 in the US, about one in twenty, while cannabis store near me holds 60,500 because Canadians use the regulated word for the regulated thing. City retail is the cluster that actually converts: weed delivery vancouver at 1,900, weed delivery toronto at 1,000, dispensary winnipeg at 390, cannabis store calgary at 320. Mail order is its own layer, with online dispensary canada at 2,900 and buy weed online canada at 1,600, which has no clean American counterpart because interstate shipping is not legal there.

One number is missing from the Canadian side on purpose. CPC is effectively zero across the whole vertical, because Google does not sell cannabis ads in Canada, so there is no auction to read a commercial signal from. Anyone pricing Canadian cannabis SEO against a paid benchmark is pricing against a market that does not exist. Difficulty still works, and it runs lower than the US on almost every product term.

The B2B cluster is the weakest part of the Canadian set and worth saying so. Cannabis seo agency at 320 and dispensary seo at 140 are real, but most of the American agency, licensing and software vocabulary lands under 50 searches a month and was cut. The category tables below this point are the US dataset. The Canadian set lives in the mapper, and the copy button exports whichever market is on screen.

Data appendix

Source: Semrush US database, accessed September 7, 2026, via phrase_fullsearch (broad match), phrase_these (batch lookup) and phrase_questions.

Metrics: Search Volume is Semrush's modeled monthly US search estimate. KD is Semrush's Keyword Difficulty Index (0-100). CPC is Semrush's average cost per click in USD; a dash means Semrush returned no CPC, which usually reflects thin advertiser data in a restricted vertical rather than zero commercial value. Intent is Semrush's classification, collapsed here into Informational, Navigational, Commercial, Transactional and Mixed.

Known limitations. Volumes are estimates that cluster onto standard buckets, so treat them as relative magnitudes rather than forecasts. Broad-match pulls in this vertical are contaminated by homonyms, and the contaminated terms are flagged in place. Canadian, UK and EU volumes are not covered here; re-pull in the relevant database before planning non-US markets. Keyword difficulty is a vendor model and should inform sequencing, never exclusion.

Total measured keywords: 1,041 unique terms across 16 categories, printed as 1,221 table rows because some keywords legitimately belong in more than one category.

RDF semantic triples

contains

1. Cannabis keyword list → contains → 1,041 measured keywords across 16 categories
2. Category 2 → contains → the near-me family led by dispensary near me at 1,830,000 searches
3. Category 12 → contains → brand terms larger than most category terms

is / has

1. Dispensary near me → is → the highest-volume keyword in the cannabis vertical
2. Sativa vs indica → is → the highest-volume comparison query in cannabis at 135,000
3. Cannabis, weed and marijuana → are → three separate search vocabularies with different intent
4. Dispensary pos → has → a $19.81 CPC against 170 monthly searches
5. Search volume → is → a modeled vendor estimate, not a count

causes / produces

1. Homonym contamination → causes → inflated volumes on weed, edibles, strain, bong, joint and grinder
2. Vocabulary mismatch between industry and consumer nouns → causes → lost rankings on vape and cart terms
3. Broad-match pulls without modifiers → produce → unusable cannabis keyword data

requires

1. Strain pages → require → first-party inventory, pricing, photography and COA data to beat strain databases
2. Legality pages → require → scheduled maintenance because state law moves
3. Hemp-derived cannabinoid keywords → require → re-pulling as legislation changes

blocks / restricts

1. Google Ads, Meta and TikTok → block → cannabis advertising
2. Advertising restrictions → make → organic search the primary acquisition channel in cannabis

outranks

1. Weedmaps and Leafly menu pages → outrank → many operators' own store pages on their own brand terms
2. Keyword variation coverage → outranks → head-term repetition in on-page correlation data

serves

1. ProStar SEO → serves → cannabis, CBD and hemp operators in the US, Canada and Europe
2. Category 15 → serves → operators rather than consumers, at low volume and high deal size
3. Category 16 → serves → existing customers searching for pricing, measurements and deals

targets

1. The near-me family → targets → local pack and store page visibility
2. The B2B category → targets → licence holders with budget rather than search volume
3. Spanish-language variants like dispensary cerca de mi → target → an underserved 18,100-search audience