We employed a nested case-control study design to evaluate the efficacy of bleach-cleaning of needles and syringes among injecting drug users (IDUs) as a means of preventing human immunodeficiency virus (HIV) infection. Sixteen HIV-seroconverters who responded to bleach use questions and who reported injecting with shared or used equipment in the 6 months prior to their first positive visit were compared with 89 controls. Controls had remained HIV-seronegative at two or more visits, reported injecting with shared or used equipment, responded to bleach-cleaning questions, and were seen at recall visits +/- 6 months from the date of seroconversion of the index case. Risk factors associated with HIV seroconversion in univariate analyses were a history of sexual intercourse with an HIV-infected partner and the frequency of speedball (mixed heroin and cocaine) injections. After adjusting for confounders, we found no evidence that bleach use protected against HIV infection.