This paper assesses policy development, service changes and trends in HIV infection and risk behaviour among injecting drug users (IDUs) in the United Kingdom. In 1986, the U.K. was faced with the possible rapid spread of HIV infection among IDUs. The combination of an outbreak of HIV infection with prevalence levels of 50% or more in Edinburgh, the recent diffusion of drug injecting, and high levels of syringe-sharing risk behaviour, suggested that HIV infection might spread rapidly through IDU populations. HIV prevention activities commenced in 1986 and developed in 1987. The first report on AIDS and Drugs Misuse by the Advisory Council on the Misuse of Drugs in 1988 was a major catalyst for change. It supported and legitimized emergent views on new ways of working with drug users. Between 1988 and 1993 innovative public health projects increased the ability to target vulnerable populations through syringe distribution, expansion of methadone treatment and outreach to hard-to-reach populations. There were major changes in service philosophy and practices, as ideas of harm minimization, accessibility, flexibility and multiple and intermediate goals were developed. There is evidence that these public health projects encouraged extensive changes in the health behaviour of IDUs. There have been major reductions in syringe-sharing risk behaviour and sharing syringes is no longer the norm. Evaluation of specific interventions (e.g. syringe-exchange) shows their importance in encouraging reductions in risk behaviour. Levels of HIV infection in IDUs remain low by international standards. Outside of London rates of about 1% have been reported; London has a low and declining prevalence of infection to around 7% in 1993; previous high levels in Edinburgh (55%) have since declined to 20%. Britain has to date avoided the rapid increase in HIV infection among injectors that has occurred in many parts of the world. The same period saw the continuation of high prevalence levels in New York and many European cities, and the explosive spread of HIV in many countries in south-east Asia. This paper acknowledges the difficulties is proving links between social interventions and epidemic prevention. It argues that there is prima facie evidence for the success of public health prevention, that the collection of intervention approaches in the U.K. had a significant impact on IDUs behaviour, and that this has helped prevent an epidemic of HIV infection among IDUs. The U.K. experience adds to the growing evidence of the significance of early interventions in encouraging behaviour change and in limiting the spread of HIV infection.