Amsterdam Coffeeshop Menu Price Report: October 2026

So what did Amsterdam coffeeshop menus actually say about prices this October? Here’s the 30-day snapshot, with the sample counts and the caveats right beside the numbers—not hidden in the small print.

The sample is recent menu evidence, not a city census

Source: menus observed on Amsterdam Coffeeshop Menus between 21 September and 21 October 2026. The snapshot was generated on 21 October 2026 and covers a rolling 30-day window; 28 of the 251 listed shops contributed fresh menus in that period.

Units: weed, hash and extracts are stated in euros per gram, computed from listings whose menus print a weight. Joints are stated in euros per listed unit. Each listing contributes its best observed price in the window, so repeated observations of that listing are not counted as separate offers. A shop can still contribute several different listings.

Sample: the archive as a whole holds 10,743 observed price lines, of which 10,690 are trusted. The fresh-window listings behind each row are printed in the table below. They are small samples, and every row should be read as a count-limited observation, not a census.

The October table

Here it is - per-gram observed prices, best observed per listing, fresh window 21 September to 21 October 2026. Sample counts included, because every number needs its receipt:

A dash means the statistic was not published in this snapshot. It is not zero and cannot be reconstructed from the median alone. Pack formats carry no published p25 here, and the joints row carries no published minimum or maximum. The report leaves those gaps visible rather than estimating them.

What the medians say

Weed single grams have a central band from €10 to €16 around a €13 median across 139 listings, with €5 and €35 as the observed tails. Weed 5g packs show a higher per-gram median at €14 across 187 listings and an upper quartile of €25. These are separate samples; the table does not establish what causes the difference or prove that matched pack offers cost more than matched loose grams.

Hash reads differently by format. Single-gram hash has a median of €15, but its p75 sits at €40, so the upper quarter of 1g listings sits far from the middle; the format ranges from €4 to €100 across 93 listings from 26 shops. Hash packs sit lower, a €14 median across 98 listings. Extracts are the most expensive category per gram by a wide margin: a €57.50 median for 1g across 40 listings from 22 shops, with pack listings at €60.

Joints: priced per unit, never per gram

The 192 joint listings from 24 shops produced a p25 of €5, a median of €7.50 and a p75 of €21 per unit. The median describes the middle observed unit listing, while the upper quartile describes a different position in that sample. Neither identifies the contents of an individual unit. A higher-priced line could need a format check, but the summary cannot establish whether it is a multipack or a special. Read the source before using it as a single-joint comparison. Without a printed weight, a unit price cannot enter a per-gram median, and this snapshot supplies no minimum or maximum for the row.

How to use this report

Read across a row before comparing down a column

Start with the format column. The weed pack median and the weed single-gram median both use euros per gram, but they describe different groups of listings. A common unit makes the arithmetic readable; it does not make the offers interchangeable. You still have to pay for the quantity in the pack, and the table does not say that a shop sells part of that pack at its calculated rate.

Then read the sample columns. Listings count menu offers, while shops count the businesses represented by those offers. A shop can contribute different products, so the listing count is not a count of independent shops. This distinction matters when you describe the result to somebody else: a middle listing price is not the price charged by the middle shop.

Finally, look at the empty cells. The hash row has no published lower quartile, so you cannot describe its complete middle-half band from this report. You can discuss its median and upper quartile separately. Keeping a sentence narrower than the table is better than filling a gap with a plausible-looking estimate.

A useful note preserves the row, not just its headline

If you save a figure for a later visit, save its category, format and observation window beside it. Copying only the median removes the information that makes the comparison possible. Copying only the minimum removes even more: you lose the distinction between an unusually low observed line and the centre of the sample.

For the weed single-gram row, the lower and upper quartiles give you a reference band. That band describes how the observed prices were distributed; it is not a recommended spending range. A line outside it is not automatically mispriced, and a line inside it is not automatically good value. Those judgments require information the table does not hold.

You can also use the table to prepare a specific counter question. If your saved line says pack, ask about that pack rather than requesting a loose gram at its converted rate. If the line says unit, confirm what the unit contains. This is a more useful outcome than treating a statistical comparison as a promise.

Comparing this report with a later snapshot

A future median could change because prices changed, because different menus entered the window, or because older menus left it. These possibilities are not separable from the headline alone. Before calling a difference a price rise or fall, compare the formats and the sample coverage in both reports.

The October report remains a dated reference even when the live comparison pages show newer observations. Its publication date should not be substituted for its observation window. When citing it, describe it as the October snapshot of observed menus, then keep the relevant row’s listing and shop counts with the figure. That leaves readers able to judge the evidence themselves.

Use the medians as anchors and the quartiles as the normal range. A line you saw at a shop sits somewhere in one of these rows; find the row of its category and format, then read where it lands. Inside p25 to p75 is common. Beyond p75 is uncommon, not impossible, and worth a second look at the menu’s date before any other conclusion. If a line falls outside the whole observed range, treat it with the same caution you would give a promise that looks too good, then verify on site. The report is 18+ information, info only, no sales.

Limitations of this report

Observed, not live: every figure is a dated menu price, and the counter price can differ. Best observed per listing: each listing contributes its best price in the window, which can place the sample below other prices printed for those listings. Sample sizes differ: joints have the largest row at 192 listings, weed packs have 187, and weed single grams have 139. Extracts single grams rest on 40 listings across 22 shops, so a change in the contributing observations can affect the quartiles. Pack rows lack a published p25 and the joints row lacks both tails; missing statistics are not estimated. Only 28 of 251 shops contributed fresh menus, so the report does not represent every listed shop. The observation window is 21 September to 21 October 2026; older archives remain historical. Always verify on site.

Where to check the current numbers

The medians and ranges update as menus are captured: weed prices and hash prices carry the per-gram tables by format, joint prices holds the per-unit list, and the cheapest ranking compares like-for-like entries. The median guide explains why the middle and not the minimum leads these reports.

How to cite a row without overstating it

Keep the observation window, category, format, listing count and shop count together whenever you quote a result. “Weed 1g median” on its own hides both the period and the sample behind it. A precise description is that the figure summarizes the best observed eligible price per listing in the stated window; it is not the average of every price a shop printed, and it is not a current quote. If a table has no quartile or range value, leave it missing rather than estimating it from a nearby category. That distinction matters most when comparing smaller groups, where a few changed observations can shift the summary. For a current visit, use the live price tables as a dated archive and confirm the menu at the counter.