The hundred men from a forest area of Ghana, without vector control or ivermectin distribution, were randomized to receive a single dose of ivermectin (150 micrograms/kg body weight) on day 1 followed by amocarzine (3 mg/kg twice daily after meals) on days 8, 9 and 10 (34 patients), the ivermectin alone (33 patients) or the amocarzine alone (33 patients). Detailed clinical and laboratory examinations were made before, during and after drug administration. On day 120, all palpable nodules were excised, fixed, sectioned, stained and examined by two blinded observers and the results compared with those for nodules from untreated controls. Mazzotti-type reactions, such as itching, rash, peripheral sensory phenomena and swellings, were more severe or frequent with amocarzine than ivermectin. Pretreatment with ivermectin markedly suppressed these reactions to amocarzine but did not affect other manifestations such as dizziness and gaze-evoked nystagmus. Ocular effects were minor in all groups. Ivermectin produced minor macrofilaricidal effects on the adult male worms, marked degeneration of intra-uterine embryos, and potent microfilaricidal effects and suppressed skin microfilariae. Amocarzine did not affect the male worms or the intra-uterine embryos, was a less potent microfilaricide and did not suppress skin microfilariae. The efficacy of ivermectin plus amocarzine was similar to that of ivermectin alone. The present results do not support the findings from the Americas and show that amocarzine has no role in the treatment of onchocerciasis in Africa.