Twenty male patients from Guatemala infected with Onchocerca volvulus received a 3 mg/kg oral dose of amocarzine twice daily for three days. The patients were randomly assigned to the sequence fasting/non-fasting and non-fasting/fasting for the morning administration on days 1 and 3. All other doses were given after food intake. Blood samples on days 1 and 3 and urine fractions from days 3 to 5 were collected for the determination of the unchanged drug and of its N-oxide metabolite, CGP 13 231. The absorption of amocarzine and CGP 13 231 was slower and sustained for longer time in fed patients than in fasting ones. The mean AUC of amocarzine was significantly higher (about 20%) in fed patients. No significant difference was found for CGP 13 231. The relative improvement of bioavailability of amocarzine due to food was less prominent than previously obtained after a high dose of 1200 mg which demonstrated a bioavailability improvement of a factor of three. Therefore, saturable dose dependent absorption processes are likely to be involved for the administration in fasting conditions. Conversely, the concentrations of amocarzine in fed patients after 150 and 1200 mg were dose proportional, thus indicating linear kinetics. The cumulative urinary excretions of CGP 6140 ranged from 0.1 to 3.8% of the daily dose. Those of CGP 13 231 ranged from 31 to 64%. This total excretion was larger than that previously recorded in fasting patients after a single oral dose. The present results confirm the improvement of the bioavailability of the drug administered after food intake.