Relative bioavailability of enteric-coated diclofenac (CAS 15307-86-5) was investigated after a single-dose administration of a multiple-unit formulation (Diclo-Puren 50, test) in comparison to a single-unit formulation (reference). The study was carried out in a four-way change-over design including a group of 12 healthy male volunteers. Each formulation was administered after 10 h of fasting or just after (5 min) finishing a meal (standard breakfast). Diclofenac plasma concentrations were measured using a selective and sensitive GLC-MS method after liquid-liquid extraction and derivatisation. Area under the curve (AUC), maximum plasma concentrations (Cmax), time of maximum plasma concentration (tmax), time of delay of first measurable concentrations (tlag) and plateau time of concentrations above minimum effective concentrations (MEC) of 50 ng/ml (tMEC(50)) and 100 ng/ml (tMEC(100)), respectively, were evaluated as pharmacokinetic characteristics. Additionally, for AUC and Cmax, 90%-confidence intervals (parametric: ANOVA, ANOVAlog, non-parametric: Mann-Whitney) were calculated to evaluate the influence of food on bioavailability of each formulation. Mean (median) relative bioavailabilities of diclofenac of the test formulation (comparison: postprandial vs. fasting conditions) were determined for the test formulation as 96% (103%) and for the reference product as 70% (83%). Mean +/- SD (median) maximum plasma concentrations of the test formulation were determined as 695 +/- 313 (677) ng/ml (fasting) and as 452 +/- 163 (456) ng/ml (postprandial). Maximum plasma concentration occurred 1.2 +/- 0.5 (1.0) h and 4.8 +/- 1.0 (5.0) h after administration, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)