1. The renal tubular excretion of cefuroxime and ceftazidime in relation to the coadministration of probenecid was investigated in eight and two healthy subjects, respectively. 2. Cefuroxime or ceftazidime were administered by i.v. infusion and 1 g probenecid was administered orally after steady state plasma concentrations of the cephalosporin were reached. 3. In a second session the same antibiotic was administered at increasing infusion rates such that three different levels of plasma drug concentration were achieved. 4. The renal clearance of antibiotic was calculated based upon unbound plasma concentration, and tubular clearance was estimated by subtracting inulin clearance from the renal clearance of the antibiotic. 5. Non-linear regression analysis was used to estimate parameters describing the saturability of tubular excretion and the effect of probenecid inhibition, i.e. EC50 and Rtub,max, could be established for cefuroxime: EC50 was 248 (s.d. 130) mg l-1 and Rtub,max was 1.852 (s.d. 0.577) mg h-1. Tubular excretion of ceftazidime was practically zero. The EC50 of probenecid for inhibition of the tubular excretion of cefuroxime was 0.80 (s.d. 0.31) mg l-1. 6. The results indicate that in the therapeutic plasma concentration range of cefuroxime its renal clearance is not saturated. Probenecid at therapeutic doses will block tubular excretion of cefuroxime almost completely.