In 1978 the third largest Swedish city, Malmö, known to have the highest suicide frequency in the country, was found to have a higher prescription rate (defined daily doses (DDD) per 1,000 inhabitants per day) of anxiolytic-hypnotic drugs (AHD) than the country, the corresponding county, other counties, and other cities, including the largest (Stockholm) and second largest (Göteborg = Gothenburg) cities. Barbiturate prescribing in Malmö was 40% higher than in Stockholm and 90% higher than in Göteborg, and the frequency of suicide due to barbiturates was three-times higher than in Göteborg. A small proportion (2.4% of all AHD-prescribing doctors) of private practitioners wrote a large percentage (24%) of all AHD prescriptions. Prescription surveillance and an information campaign in Malmö were accompanied by a 4-year decrease in AHD prescribing (12%), in AHD abuse (40%), in barbiturate prescribing (45%), and in barbiturate suicides (70%). The total suicide rate was reduced by 25%. There was no corresponding 4-year increase in suicide due to other drugs, or by other means, but after 5 to 7 years there was an increase in suicide by non-pharmacological means. The contribution of benzodiazepines to the frequency of suicide was very small, whereas their contribution to AHD abuse was considerable. In Göteborg, where no corresponding intervention was carried out, there was also a reduction in barbiturate prescribing (34%) and in barbiturate suicides (45%), but in contrast there was a continuous increase both in overall AHD and benzodiazepine prescribing, surpassing Malmö after 5 years. Far from a reduction there was a 7-year increase in the overall frequency of suicide. Apparently, AHD abuse and suicide can be greatly reduced by restricted prescribing of AHD, and this may but need not be accompanied by an increase in suicide by other means. Targeted drug information campaigns may assist in changing prescription patterns and their medical and social impact.