We used a prospective, open-label randomized trial to evaluate two treatment regimens in Tanzanian children two months to four years of age presenting to a hospital with a packed cell volume (PCV) < 25%. Treatment was either standard (14 days of ferrous sulfate and an antimalarial) or extended (three months of ferrous sulfate and three antimalarial treatments). The prevalence of anemia was measured two weeks after completion of treatment and six months after recruitment. Two weeks after completing treatment, the prevalence of PCV < 33% was 58% in the standard treatment arm and 44% in the extended treatment group (P = 0.04), and the mean PCV was significantly higher in the extended treatment arm (32.1%, SD = 4.5% versus 30.8%, SD = 4.9%; P = 0.031). However, there was no difference in the prevalence of PCV < 25% in the first survey, and the benefits of extended therapy were only apparent six months after recruitment in children compliant with the extended treatment (odds ratio of PCV < 25% = 0.16, P = 0.06). Compliance was satisfactory in only 39% (82 of 209) of the children in the first week of treatment. Extending the duration of therapy and improving compliance may have health benefits for anemic children in malaria-endemic settings.