Pharmacokinetic (PK) parameters for terbina®ne were assessed in 15 elderly and 15 young healthy subjects randomized to receive 250 mg Lamisil 1 once daily for 15 d in a two-period, two-treatment, two-sequence, crossover (fed versus fasted) design within age groups. On each treatment day except days 8 and 15, subjects took Lamisil 1 with food at 8:00 a.m. On days 8 and 15, subjects took the drug under either fed or fasting conditions according to treatment sequence, and 24 h PK pro®les were obtained. Two analyses of the pharmacokinetic data were undertaken. In a noncompartmental analysis, AUC 0±24 h , C max , C 0 h , and t max were computed for each subject on each of days 8 and 15, and the in¯uences of food condition and age on these variables were assessed by analysis of variance. AUC 0±24 h and C 0 h were found to be larger (p<0´5) among elderly subjects than young subjects on day 15, and t max was prolonged (p<0´05) under the fed condition on day 15. Similar trends on day 8, as well as generally higher exposures under the fed condition on both days, were not statistically signi®cant. A three-compartment model ®tted to the complete sequence of 33 terbina®ne concentrations measured over 29 d for each subject separately permitted a withinsubject assessment of the food eect that con®rmed prolonged absorption and increased bioavailability (p<0´05) under the fed condition. Across-subject comparisons of oral clearances estimated with the model also con®rmed the increased exposures (lower clearances) among the elderly (p<0´05). The drug was well tolerated in both age groups. In particular, greater drug exposure in the elderly did not result in greater toxicity, as indicated by the safety evaluations in the study. &1997 by John Wiley & Sons, Ltd.