Pilot studies of the safety and efficacy of 3 drugs thought to have anticryptosporidial activity were carried out to determine whether any of them are suitable for large-scale clinical trials. Open studies of the use of azithromycin, letrazuril and paromomycin in patients with acquired immunodeficiency syndrome (AIDS) and confirmed cryptosporidial diarrhoea for at least a month. Azithromycin 500 mg daily was ineffective. Letrazuril 150-200 mg daily was associated with an improvement in symptoms in 40% of patients treated and cessation of excretion of cryptosporidial oocysts in the stool in 70%; however biopsies remained positive. Paromomycin therapy was associated with a complete resolution of diarrhoea in 60% of patients treated and some improvement in symptoms in a further 5% but it did not eliminate the infection. None of the drugs had any major toxicities. Dose escalation studies of azithromycin should be performed. Letrazuril should be further investigated for efficacy in double-blind placebo-controlled trials. Paromomycin appears to result in prolonged symptomatic remission of cryptosporidial diarrhoea, but has no effect on cryptosporidial cholangitis.