OBJECTIVE: While data on drug abuse liability have been determined for selected drugs in experimental studies, there is limited information available about drug abuse in real life. This study presents the results of a simple experimental epidemiologic survey of prescription forgeries in community pharmacies in the Midi-Pyrenees area (southwest France). During four periods (September-October 1991, January-June 1992, September-October 1992 and January-June 1993), resident pharmacy students in several volunteer pharmacies were asked to complete a specific report form for each suspect prescription request. The main criteria used to identify forgeries were: inadequate dosage, multiple use of the prescription form, drafting not in accordance with the rules of prescription or false prescription forms (stolen prescription forms, photocopies). RESULTS: A total of 165 falsified prescriptions were collected. The 305 drugs involved in these forged prescriptions were opiate analgesics, benzodiazepines, amphetamines and minor opiate analgesics. Medications were essentially buprenorphine, flunitrazepam (in 2 mg dosage), phenobarbitone+amphetamine and chlorazepate. CONCLUSION: These results suggest that forged prescriptions can be used as an indicator of potential abuse liability of marketed psychoactive drugs. Although this survey cannot describe the real prevalence of the misuse or abuse of drugs, it constitutes a useful warning epidemiologic system to elicit early observations regarding new misuses of drugs as they are requested at the pharmacy.