We investigated the benefit of treating HIV-1-infected children with monotherapy where resources are limited. A retrospective chart review was undertaken in 12 symptomatic HIV-1-infected children treated with zidovudine or didanosine for at least 2 months. The main outcome measure was the effect on hospitalization. Anti-retroviral therapy was commenced in nine children because of prolonged or frequent hospitalization. Of three whose primary indication was bleeding secondary to thrombocytopenia, two had been hospitalized owing to severe intercurrent illness. One child had failure to thrive and another encephalopathy. Monotherapy was considered beneficial in all cases. Median duration of follow-up was 6.5 (2-31) months. The hospitalization index (days in hospital before and after start of monotherapy, divided by the total number of days before and after start of monotherapy) decreased from a median of 0.115 prior to therapy to 0.037 on therapy (p = 0.045, Wilcoxon matched pairs test). This study presents observational data supporting the investigation of monotherapy in resource-poor countries. It was associated with a significant reduction in hospitalization and appeared to result in clinical improvement. Prolonged or frequent hospitalization might represent a novel indication for use because in our setting the cost of hospitalization could potentially provide a 12-month-old infant with monotherapy for 2 months.