A rapid assessment of drug abuse in Nepal was conducted at different sites, including eight municipalities in the five development regions of the country. To interview various groups of key informants, such methods as semi-structured interviews, in-depth interviews and focus group discussions were used. A snowball sampling strategy for respondents who were drug abusers and a judgemental sampling strategy for the non-drug-using key informants were applied. About one fifth of the sample was recruited from the treatment centres and the rest from the community. Drug abusers in prison were interviewed, and secondary data from treatment centres and prisons analysed. The study revealed that the sample of drug abusers had a mean age of 23.8 years and was overwhelmingly male. Most respondents lived with their families and were either unemployed or students. About 30 per cent of the sample was married. A large majority of the sample had a family member or a close relative outside the immediate family who smoked or drank alcohol and a friend who smoked, drank or used illicit drugs. Apart from tobacco and alcohol, the major drugs of abuse were cannabis, codeine-containing cough syrup, nitrazepam tablets, buprenor-phine injections and heroin (usually smoked, rarely injected). The commonest sources of drugs were other drug-using friends, cross-border supplies from India or medicine shops. The commonest source of drug money was the family. There has been a clear trend towards the injection of buprenorphine by abusers who smoke heroin or drink codeine cough syrup. The reasons cited for switching to injections were the unavailability and rising cost of non-injectable drugs and the easy availability and relative cheapness of injectables. About a half of the injecting drug users (IDUs) commonly reported sharing injecting equipment inadequately cleaned with water. Over a half of IDUs reported visiting needle-exchange programmes at two of the study sites where such programmes were available. Infection by the human immunodeficiency virus (HIV) appears to be low among IDUs, although systematic surveillance is absent. Two thirds of the sample had experienced sexual intercourse. The last sex partners reported by respondents were commercial sex workers, wives or girl friends. Condom use was low with primary partners and relatively high with sex workers. Treatment facilities, mostly located in the central urban areas of the country, are meagre. An overwhelming majority of drug abusers felt the need to stop abusing drugs. Cost-effective drug treatment and HIV prevention programmes for IDUs are urgently needed in all areas of the country.