The aim of this study was to evaluate the efficacy of cefaclor in the prophylaxis of recurrent acute otitis media (AOM) in human immunodeficiency virus (HIV)-infected children. The study was carried out in children born between 1 January 1986 and 31 December 1996 who had been vertically HIV infected Patients who had experienced recurrent AOM between October 1997 and March 1998 (period 1) were eligible for the trial. Recurrent AOM was defined as the occurrence in the same patient of three or more episodes of AOM within 6 months of the observation period. Patients recruited for this trial received cefaclor at a dose of 20 mg/kg once daily for 6 months between April and September 1998 (period 2). Clinical observation was carried out in periods 1 and 2 and for the first 6 months after prophylaxis, i.e. October 1998 - March 1999 (period 3). Natural killer-cell activity, phagocytosis and myeloperoxidase activity were determined before and at the end of the prophylactic period. For each period, CD4-cell count measurement and CD4-positive cell class were recorded. Seventeen children were recruited for this trial. No significant differences were observed in natural killer-cell activity between periods 1 and 2, nor were any significant differences observed in CD4-positive cell class or CD4-positive cell count between the three periods. However, cefaclor administration was associated with a reduction in the number of AOM episodes in 100% of cases and a mean increase in myeloperoxidase activity in 57% of cases. This suggests that cefaclor may be useful in the prophylaxis of recurrent AOM in HIV-infected children.