This open, randomized, single-dose crossover study investigated effects of a high-fat meal on the pharmacokinetics of roflumilast and its major active N-oxide metabolite. Twelve healthy subjects received oral roflumilast 500 microg (2 x 250 microg) after overnight fasting and after breakfast. Blood was sampled up to 54 hours for pharmacokinetic profiling of roflumilast and N-oxide. Geometric mean ratios (fed/fasted) for point estimates (PE) and 90% confidence intervals (CI) were calculated for AUC(0-last), AUC(0-infinity), and C(max) of both compounds. After the meal, roflumilast C(max) (PE, 0.59; 90% CI, 0.49-0.70) was modestly reduced; N-oxide C(max) (PE, 0.95; 90% CI, 0.90-1.01) was unchanged. Roflumilast t(max) was delayed in fed state (2.0 +/- 0.4 hours) versus fasted state (1.0 +/- 0.2 hours); N-oxide t(max) was unaltered. No significant food effect on roflumilast AUC(0-last) (PE, 1.04; 90% CI, 0.90-1.21), AUC(0-infinity) (PE, 1.12; 90% CI, 1.00-1.25), and respective N-oxide AUCs (PE, 0.91; 90% CI, 0.79-1.04; PE, 0.99; 90% CI, 0.92-1.06) occurred. Because roflumilast N-oxide is the major contributor to roflumilast's overall pharmacologic effects, these findings suggest that roflumilast can be taken with or without food.