Swiss drug policy aims to reduce recreational drug consumption and its negative consequences. The strategy is based on a four pillars concept: prevention, therapy, harm reduction and repression. Among harm reduction programs drug checking (DC) facilities have emerged, allowing drug users to check the presence of adulterants or other non-expected substances, and to gain information concerning the purity of the product, without encouraging drug consumption. In 2019, one DC opened in Geneva (Switzerland) after receiving an authorization from the Federal Office of Public Health, and the results of 3 years of activity are presented. Drug users brought samples anonymously to DC. Samples were send to the laboratory for analyses, and results are transmitted to DC, where drug users could obtained and discussed the results. Samples were dissolved in methanol, and a general unknown screening was performed by using gas chromatography coupled to mass spectrometry (GC-MS) directly and after acetylation. Depending of the substances, quantitative analyses were performed by using GC-MS, liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS), or liquid chromatography coupled to diode array detector (LC-DAD). A total of 883 samples were analyzed from June 2019 to April 2022. The majority of samples were powders (67%), tablets (11%), plants (8%), and blotting papers (6%). For 71 samples (8%) the detected substances did not correspond to those announced by consumers, and in 23 samples (3%), no psychoactive substances were detected. Cocaine was detected in 290 samples (33%), with a median purity of 75%. In these samples, 11% contained also levamisole, 11% caffeine, and 7% phenacetin. Sympathomimetics were detected in 278 samples (31%). MDMA (16%), amphetamine (4.8%), 3-MMC (3.3%), 2C-B (2.6%), and 4-MMC (1.4%) were the most detected substances in this group of substances. MDMA was detected in powder specimen with a median purity of 85%, and in tablet form with a median amount of 133 mg (range: 3–320 mg). More rarely, methamphetamine, MDEA, 6-APB, 2C-D, 2C-E, BOD, 2-Br-4,5-DMPEA, DOB, DOC, DOM, MBDB, 3-FA, 5-MMPA, 5-MAPB, and 4-CMC were observed. Cannabinoids were detected in 84 samples (9.5%), and among these samples, 11 contained synthetic cannabinoids [5-Fluoro-MDMB-PINACA (3 cases); MDMB-4en-PINACA (8 cases)]. LSD and other hallucinogenic substances (1cP-LSD, psilocin, DMT, 4-ACO-DMT, 4-HO-MET) were detected in 85 samples (9.6%). Ketamine was detected in 68 cases (7.7%), with a median purity of 76%. Heroin was detected in 64 cases (7.2%), with a median purity of 20%. Benzodiazepines (alprazolam or etizolam) were detected in 7 cases (0.8%). The DC activity reached out to consumers who were not yet connected to prevention structures. The prevalence of the detected substances in the DC samples analyzed confirmed roughly what was known about the local market for illicit substances. However, some unexpected results were observed, such as the number of samples containing ketamine, or some new psychoactive substances. Interestingly, a couple of months after the beginning of COVID pandemic, 2 synthetic cannabinoids never detected before in Western Switzerland were observed, showing the interest of the DC facility in a harm reduction strategy, and more generally for public health prevention. Knowledge concerning the market of illicit psychoactive substances is complex and often difficult to assess. In this context, the analyses carried out for the Geneva DC have demonstrated a definite interest in better knowing and understanding the dynamics of the illicit substance market, complementing other information from police and customs seizures, wastewater analyses, used needle analyses, suspected DUID case analyses, clinical toxicology case analyses and postmortem case analyses.