1 The aim of this study was to establish whether the renal tubular excretion of benzylpenicillin is saturable and whether the effect of probenecid on the tubular excretion of benzylpenicillin is dose-dependent. 2 Each of four volunteers underwent three experiments. In each experiment benzylpenicillin was administered by continuous infusion, such that three different consecutive concentration levels were reached. In the first experiment no probenecid was given; in the second and third experiments, probenecid was administered by continuous infusion at a low and higher rate, respectively. 3 Plasma and urinary concentrations of benzylpenicillin were determined at 30 min intervals by high performance liquid chromatography. 4 By fitting the equation Rtub = Rtub,max.Cp/(EC50 + Cp) to the values of the tubular excretion rate found for benzylpenicillin (Rtub) vs the free plasma concentration (Cp), the values of Rtub,max and EC50 could be calculated: 3350 (+/- 606) mg h-1 for Rtub,max and 48.0 (+/- 17.8) mg l-1 for EC50 (in the absence of probenecid). 5 The EC50 for benzylpenicillin increased significantly with increasing doses of probenecid. 6 The dose of probenecid at which 50% of the excretory system is occupied by probenecid in the absence of benzylpenicillin (ED50) ranged from 13.2 to 108.5 mg h-1. 7 The EC50 of probenecid in one subject could actually be measured: 52.3 mg l-1. 8 Extrapolating these results to the clinical situation, the commonly used daily dose of 2 g of probenecid is likely to be close to the maximal effective dose for inhibition of the tubular excretion of benzylpenicillin.