The nonsteroidal antiinflammatory drugs (NSAIDs) oxaprozin and piroxicam have long elimination half-lives (t 1/2 approximately 55 hours), permitting once-daily dose regimens. The protein-binding characteristics of these drugs, however, vary widely. This study examines the effect of these binding differences on the drugs' disposition kinetics at steady state. A total of 52 participants (26 young healthy volunteers, and 26 elderly osteoarthritic patients, 15 men and 37 women (2 of them poor metabolizers of debrisoquine [CYP2D6]) completed the two-period, two-treatment, randomized, single-dose and 21-day, once-daily multiple-dose, cross-over study. Doses of oxaprozin and piroxicam were 1,200 mg once daily and 20 mg once daily, respectively. Mean single-dose kinetic parameters of oxaprozin versus piroxicam did not differ more than +/-14% (t1/2, 53.0 versus 57.4 hours; apparent oral clearance adjusted for 70-kg body weight [Clpo], 0.139 versus 0.121 L/hr; apparent volume of distribution adjusted for 70-kg body weight [Vd/F]; 10.2 L versus 9.13 L). Protein binding was plasma-concentration dependent with oxaprozin (range, 10-400 mg/L) but not with piroxicam (range, 1-30 mg/ L). Steady-state conditions were established within 3 days with oxaprozin but took almost 12 days with piroxicam. Compared with the single-dose values, steady-state Clpo (Clpo,ss) and Vd/F of total drug increased with oxaprozin by almost 127% but remained within +/-10% with piroxicam. Post-steady-state apparent t 1/2 of the total and unbound drugs of approximately 62 hours were similarly prolonged with piroxicam but differed substantially with oxaprozin (50.6 hours [total drug] versus 23.8 hours [unbound drug]). Single dose Clpo (Clpo,sd) values of both NSAIDs were significantly correlated in the study populations. With both NSAIDs, Clpo in the two poor metabolizers of debrisoquine was within +/-20% of mean values for the population. Clinically important age- and gender-dependent decreases were not observed in the weight-adjusted, Clpo,sd or Vd/F values of the total drug for either NSAID. Clearances of the two NSAIDs were significantly correlated, suggesting that a common P450 isozyme (most likely CYP2C9, in that piroxicam is a known substrate of this isozyme) may be at least partly involved in the oxidative metabolism of these NSAIDs.