BackgroundGiven the growing worldwide market of non-prescription drugs, monitoring their misuse in the context of self-medication represents a particular challenge in Public Health. The aim of this study was to investigate the prevalence of misuse, abuse, and dependence on non-prescription psychoactive drugs.MethodDuring one month, in randomly solicited community pharmacies, an anonymous questionnaire was offered to adults requesting paracetamol (control group), codeine combined with paracetamol in analgesics, or sedative H1 antihistamines. Responses about misuse (drug use not in agreement with the Patient Information Leaflet) abuse (excessive drug use having detrimental consequences), and dependence (established according to questions adapted from the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition criteria) on psychoactive drugs were compared to those of the paracetamol control group.Results295 patients (mean age 48.5 years, 68.5% of women) having used one of the studied drugs during the previous month were included. Misuse and dependence to codeine analgesics concerned 6.8% and 17.8% of the patients exposed to these drugs, respectively, (n = 118), which was significantly higher than for paracetamol. 19.5% had used codeine analgesics daily for more than six months. Headache was the most frequent reason for persistent daily use. A high prevalence of persistent daily users of sedative H1 antihistamines was also observed. Whereas these drugs are recommended only for short treatment courses of occasional insomnia, 72.2% of the participants having taken doxylamine (n = 36) were daily users, predominantly for more than six months.ConclusionsResults on misuse and dependence on non-prescription codeine analgesics suggest that chronic pain, in particular chronic cephalalgia, requires better medical care. In addition, as for hypnotics on prescription, persistent use of doxylamine for self-medication is not justified until an acceptable benefit-risk ratio for chronic sleep disturbance is shown by clinical data.