Scottish sampling in the latest phase 3 release covers August 2024–October 2025. Home Office funding for Scottish sites was withdrawn in late 2025. In a response dated 29 June 2026 (published 13 August), the Scottish Government said it knew of no operational illicit-drug wastewater testing in Scotland at that time.
Home Office WAND · published 14 October 2026
Wastewater analysis
Drug residues across sampled wastewater systems provide a separate signal from surveys, treatment and policing. Consumption estimates add model assumptions.
Editorial review: 4 November 2026Comparable English trend16 sitesApproximately 22% of England · updated population estimates
Phase 3 network52 sitesEngland and Scotland · approximately 31% of England
Additional phase 3 analytes18Detection findings · not consumption estimates
Current phase 416 sitesEngland · quarterly since April 2026
Wales and Northern Ireland were outside this WAND programme. Wales had a separate programme assessing narcotic-drug detection at commissioning. This dashboard has no verified comparable public consumption series for either nation, so no estimate is presented.
National estimates
Latest published annual England consumption estimates — August 2024 to July 2025
Historical annual national series · estimated street-drug mass (kg) · not 2026 national totals
These annual estimates use the earlier model extrapolating from 42 English sites (28% coverage in that publication), revised 8 October 2026. They are separate from the newer 16-site trend series. Bars start at zero. Mass does not rank the number of people using drugs, doses or relative harm. Colour identifies the substance.
Trend signal
Change from 2021 to 2026
Published percentage changes · comparable 16 English sites · January–April sampling windows
Cocaine+17%January–April 2021 to 2026 · latest 2025–26: -7%
Ketamine+194%January–April 2021 to 2026 · latest 2025–26: -11%
Heroin-49%January–April 2021 to 2026 · latest 2025–26: -14%
Amphetamine-55%January–April 2021 to 2026 · latest 2025–26: -39%
MDMA+248%January–April 2021 to 2026 · latest 2025–26: +5%
Methamphetamine+74%January–April 2021 to 2026 · latest 2025–26: +8%
Largest estimated mass; latest fall
Cocaine remains the most consumed of the six monitored drugs by estimated mass. The latest comparison fell 7%; this does not establish lower harm.
−11% in the latest comparison
The early 2026 estimate fell after sustained growth and remains 194% above 2021. Treatment demand and bladder harms require separate evidence.
Methamphetamine +8%; amphetamine −39%
These are changes in modelled consumption at sampled English catchments. They do not describe the entire UK population or identify numbers of users.
Keep the uncertainty visible
MDMA rose 5%, with a published 2026 standard error of ±201 mg. Heroin fell 14%, but its unstable biomarker makes interpretation particularly uncertain. No significance test is claimed.
Consumption and harm are different measures
Decreases in estimated mass do not establish reductions in dependence, drug-related harm, hospital presentations or mortality. These trend estimates describe the sampled catchments, not the whole UK.
02
Geographical pattern
Areas repeatedly identified as higher or lower within sampled WAND locations
England and Scotland
Reported high and low consumption signals
England: August 2024–March 2026 · Scotland: August 2024–October 2025 · sampled sites only
CocaineNorth East, Scotland, South West (specific site)East Midlands, Yorkshire and the Humber, West Midlands (specific site)
KetamineEast of England, North West, North East (specific site)North East (specific site), Scotland, West Midlands
HeroinNorth East, Scotland, London (specific site)Not detected consistently enough to identify lower-consumption areas
AmphetamineEast of England, North West, Yorkshire and the Humber (specific site)East of England (specific site), West Midlands, South West
MDMANorth West, London, South East (specific site)East Midlands, West Midlands
MethamphetamineLondon, South East, Yorkshire and the Humber (specific site)No region identified as notably lower
Site qualifier
“(specific site)” qualifies a single sampled treatment plant; an unqualified region indicates a more consistent pattern across several sampled sites. Different sites explain why the North East and East of England can appear in both columns. These catchments cannot establish definitive rankings for whole regions.
03
Weekly pattern and method
Weekly patterns below describe August 2024–July 2025 evidence; the current programme is shown separately.
MDMA and cocaine
Higher estimated consumption at weekends and lower levels on weekdays. MDMA has a Monday–Tuesday tail because excretion can take one to two days.
Ketamine, amphetamine and methamphetamine
These substances showed a steadier aggregate weekly pattern. This does not establish how often any individual uses them.
Heroin
The 6-MAM biomarker is unstable, so low estimates and day-to-day patterns require particular caution.
Phase 3 windows; phase 4 quarterly
Historically, phase 3 collected seven consecutive days of 24-hour composite samples per site, approximately every two months. Phase 4 began in April 2026 with quarterly sampling at 16 English sites.
Current programme — phase 4 from April 2026
Quarterly sampling at 16 English sites monitors cocaine, ketamine, amphetamine, methamphetamine and heroin. Presence/absence testing covers 5-aminoisotonitazene and mephedrone (4-MMC). MDMA and the other routine phase 3 substances are no longer included; MDMA’s historical 2021–2026 series remains available below. Historical Scottish findings do not represent a currently operational WAND programme.
Home Office programme and methods ↗Take 24-hour composite samples at wastewater treatment plants.
Test for drugs and human metabolites such as benzoylecgonine and norketamine.
Adjust for flow, metabolism, deposition and retail purity.
Extrapolate England estimates with uncertainty intervals; Scotland is not nationally extrapolated.
Experimental
Use with care
Open WAND collection ↗WAND cannot identify how many people used a drug, who used it, individual purity, or whether every detected amount was consumed. Cannabis is not measured. Limited nitazene detection does not demonstrate absence of use. Wales and Northern Ireland were outside this programme, and national extrapolation currently applies to England only.
Six years of comparable wastewater estimates
The same 16 catchments, covering approximately 22% of England's population after updated population estimates, allow a more consistent view of change than comparing different monitoring networks.
Colour identifies the substance; it is not a harm or confidence score.
Ketamine · 2026 mean953 ± 38 SEM
Estimated street-drug mg / 1,000 people / day
2021 → 2026+194%
Published change in comparable catchments
2025 → 2026−11%
January–April windows compared
Points = estimated means. Vertical bars = ±1 SEM. On a narrow screen, scroll the chart to see every year.
| January–April | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|---|
| Cocaine | 3,206 ± 109 | 4,164 ± 62 | 4,328 ± 156 | 4,975 ± 201 | 4,045 ± 98 | 3,743 ± 134 |
| Ketamine | 324 ± 8 | 353 ± 5 | 391 ± 15 | 691 ± 34 | 1,066 ± 30 | 953 ± 38 |
| Heroin | 1,921 ± 267 | 1,138 ± 104 | 1,289 ± 72 | 1,421 ± 204 | 1,154 ± 271 | 989 ± 74 |
| Amphetamine | 2,711 ± 144 | 2,167 ± 111 | 3,978 ± 333 | 3,311 ± 833 | 1,978 ± 111 | 1,208 ± 39 |
| MDMA | 93 ± 7 | 139 ± 7 | 158 ± 11 | 257 ± 25 | 311 ± 28 | 325 ± 201 |
| Methamphetamine | 149 ± 6 | 135 ± 4 | 157 ± 5 | 192 ± 8 | 240 ± 8 | 260 ± 12 |
Read the measure correctly
These are population-weighted, retail-purity-adjusted estimates of street-drug mass, rather than pure-drug mass, residue loads or the percentage of people using drugs. Sampling windows are January–April; the source describes seven-day samples, rather than continuous monitoring. SEM describes uncertainty around the estimated mean. It is not a confidence interval and does not capture all systematic uncertainty from excretion, disposal, sewer losses or the model.
Different substances use their own zero-based vertical scale when selected. Published percentage changes use the source's estimates; rounded table values may not reproduce those percentages exactly. The October 2026 release extends the revised series, which superseded the March 2025 report. Percentage changes alone do not establish statistical significance. The published 2026 MDMA standard error is ±201; it is retained without smoothing or correction. Compare this signal with the ketamine UK profile and drug prevalence and treatment evidence.
Additional drug detections: what the new evidence reveals
England: August 2024–March 2026. Scotland: August 2024–October 2025. Detections describe sampled wastewater; they cannot quantify consumption, prevalence, population exposure or market share.
Sporadic signals need cautious interpretation
5-aminoisotonitazene, a nitazene analyte described as a metabolite, was detected across the reported regions over two cycles: July–August and August–October 2025. An analyte detection does not establish which parent product was consumed. Isotonitazene was also reported, but its regional location is disputed between the source files. Norfentanyl, a fentanyl metabolite, appeared in six English regions during July 2025–January 2026; medical use can contribute.
Most tested nitazene-related analytes were not detected: 4-hydroxynitazene, metodesnitazene, N-desethyl metonitazene, N-desethyl protonitazene, N-pyrrolidino etonitazene and N-pyrrolidino protonitazene. This does not establish their absence from supply.
Pregabalin and gabapentin across all reported regions
Both were detected in all ten reported regions during March–October 2025, and again in the nine English regions in January–March 2026. Wastewater cannot distinguish prescribed medical use from non-medical consumption or identify combinations used by individuals.
Separate drugs from their metabolites
Temazepam and nordiazepam were detected across all reported regions. Diazepam was reported in six regions. Nordiazepam is a metabolite associated with diazepam and other medicines. Alpha-hydroxyalprazolam was detected in Yorkshire and the Humber and alpha-hydroxybromazolam in Scotland; alpha-hydroxyetizolam was not detected when tested. These metabolites do not establish illicit supply or identify a product’s contents.
A limited South West detection
The report and spreadsheet identify a South West detection in the May–July 2025 testing cycle. Xylazine was not detected in the other tested region-cycle cells. Low frequency does not establish absence from drug supply: concentrations, sampling windows and analytical limits can prevent detection.
Testing coverage and a source discrepancy
The 18 additional analytes were not tested in every cycle. In the data tables, D means detected, ND means tested but not detected, and NT means not tested. These categories must not be combined. Scotland was not tested in the later cycles after October 2025.
For isotonitazene, report Table 3 names the North West, while both the spreadsheet’s Overall sheet and isotonitazene worksheet place the October–December 2024 detection in Yorkshire and the Humber. The location remains unresolved as at 4 November 2026; neither is used here as a confirmed regional finding.
What wastewater can — and cannot — tell us
A laboratory detects specific chemicals in a catchment. Further assumptions turn those signals into consumption estimates.
Benzoylecgonine (BZE), cocaine and cocaethylene
BZE supports evidence of cocaine consumption. Cocaethylene is associated with cocaine and alcohol consumed together. These aggregate measurements do not distinguish the number of users or their ages.
Ketamine and norketamine
The parent–metabolite relationship helps assess direct disposal. Human medical and recreational use can both contribute; the pair does not identify intent. Veterinary use may also affect the ketamine signal.
6-monoacetylmorphine (6-MAM)
6-MAM is unstable and inconsistently detected. These experimental estimates need particular caution. Low or absent detection cannot establish low heroin consumption or a reliable ranking of areas.
Amphetamine · parent compound
The signal can include amphetamine use, methamphetamine metabolism and prescribed dexamfetamine or lisdexamfetamine. Aggregate WAND data do not quantify these contributions. Direct disposal is another uncertainty.
MDMA · parent compound
Parent-only measurement cannot fully separate consumption from disposal. Weekend signals may extend into Monday and Tuesday because excretion takes time; a sample day is not necessarily the day of use.
Methamphetamine · parent compound
Parent-only measurement is vulnerable to disposal. Methamphetamine can be metabolised to amphetamine, so the two wastewater signals are not wholly independent measures of separate drug use.
Cannabis is outside the WAND panel
WAND does not currently test for cannabis because of sampling complexity. EUDA monitors the metabolite THC-COOH, but partitioning between liquid and suspended solids affects measurement. A missing WAND value is not zero cannabis use, and European residues cannot supply a UK national estimate.
Prevalence and load are different measures
A survey estimates how many people report use within a defined population and period. Wastewater measures a catchment signal and can support estimates of aggregate mass. More mass could reflect changes in the number of users, quantity per user, visitors or other inputs. A small group with frequent or high-quantity use can influence the total. Surveys have different sampling frames and recall periods, so their trends need not match wastewater. It cannot identify users, age groups, dependence or individual risk.
Amphetamine, methamphetamine and prescribed stimulants
Controlled human studies establish that methamphetamine can produce urinary amphetamine, and that lisdexamfetamine is converted to d-amphetamine. Prescribed dexamfetamine can also contribute. Methylphenidate is measured through different biomarkers; it is not interchangeable with amphetamine.
These pathways are established. Their relative contribution to the displayed English catchment signal is unquantified. The Home Office says comparison with NHS prescribing and dispensing data needs further research. No percentage here is attributed to prescribed or illicit sources.
Medical use, disposal, sewers and population denominators
Medical amphetamine and diamorphine are included in aggregate signals. Ketamine can enter wastewater through human medical use and veterinary sources. Measuring ketamine with norketamine supports interpretation of disposal, but does not distinguish medical from recreational human use.
Large deposition events can be excluded; regular smaller disposal may remain. Parent-only stimulant markers cannot reliably distinguish use from disposal. Biomarker breakdown, sewer travel, rainfall, flow measurement and losses add uncertainty. Residents used in population denominators do not fully capture commuters and visitors. Changes in site coverage or sampling methods can create apparent trends, so compare like with like.
Confidence depends on the finding
These editorial bands apply to the named claim. They are not probabilities, statistical confidence intervals, clinical risk scores or a ranking of substances.
What WAND measured
The official report identifies analytes, sampled catchments and reporting periods. That provides strong provenance for the published signals within the monitoring network.
Direct analytical evidence · sampled geography onlyExperimental consumption estimates
A consistent 16-site series supports cautious interpretation of aggregate change. Consumption is back-calculated, with uncertainty about excretion, disposal and sampling. Heroin needs additional caution.
Modelled measure · England catchments, January–April 2021–2026Established metabolic pathways
The cited controlled human studies observed amphetamine after methamphetamine or lisdexamfetamine administration. These peer-reviewed findings support a pathway, while its share of the English wastewater signal remains unquantified.
Peer-reviewed interpretation · human metabolism, not UK source attributionWho used drugs, and why
These aggregate data cannot determine youth prevalence, user counts or dependence, or quantify prescribed versus illicit contributions. Low detection is not proof of absence.
Evidence gap · additional sources requiredKeep geographical coverage explicit
The comparable-site chart is not a national total. The separate experimental England annual model uses 42 sites, covering 28% of the population, for August 2024–July 2025. Its uncertainty includes additional model and extrapolation uncertainty, which is different from the SEM in this chart. Historical Scottish coverage and England's model must not be presented as a current UK-wide monitoring feed.
Implications for drug intelligence and public health
Use catchment signals to guide further investigation and service planning alongside other evidence.
Triangulate a change
Compare like-for-like sampling periods with treatment presentations, surveys, hospital and emergency activity, toxicology, drug checking, police seizures and local intelligence. Agreement strengthens a constrained assessment; disagreement may reflect different populations or measures.
Investigate ketamine needs
The 11% fall in early 2026 changes the short-term trajectory, while estimates remain 194% above 2021. Compare this with treatment demand, bladder harms and hospital presentations before changing local capacity. Wastewater alone cannot identify age, dependence or unmet need.
Check attribution and coverage
Ask whether disposal, prescribing, veterinary inputs, population movement or sampling changes could affect the signal before attributing it to an illicit market. Keep confirmed official data and TD frontline estimates visibly labelled.
Plan without false precision
Cocaine’s 7% decline does not establish lower cardiovascular harm, dependence, hospital activity or mortality. Highly potent synthetic opioids may evade wastewater detection at very small quantities. Low detection must not be read as low overdose risk or used to reduce preparedness.
European comparison and the UK city-data gap
EUDA / SCORE provides useful European context using population-normalised residue loads, generally from a seven-day campaign. Those values are a different metric from the WAND consumption estimates shown here. They should not share a numerical scale.
Named UK city-year source rows could not be independently verified from the current downloadable dataset during this review. A UK city comparison is therefore omitted. The gap does not imply that no monitoring took place.