10% of young male healthy volunteers have a total bilirubin value over 20 mumol/l; thus such a value appears not relevant as screening cut off point in clinical pharmacology. This study was intended to confirm if a 27 mumol/l cut off point previously defined by the authors does not support a risk. This study dealt with 487 subjects who had together measurements of total bilirubin value and lab. tests of liver cytolysis, cholestasis or hemolysis during the selection process. 48 subjects (9.8%) had a total bilirubin value over 20 mumol/l. Correlation tests do not provide arguments of cytolysis, cholestasis or hemolysis and there was no argument in favor of Gilbert's syndrome. Out of 48 hyperbilirubinemic subjects only 22 were included in clinical pharmacology studies. In more than 60%, the total bilirubin value returned to normal spontaneously and in no case appeared a significant clinical, biological, pharmacokinetic or dynamic abnormality. Except a possible increase of slow acetylor frequency, the medical literature analysis does not show any relevant modification in metabolism, pharmacokinetics or pharmacodynamics until a 40 mumol/l value of total bilirubin. Thus, the 27 mumol/l value of total bilirubin previously proposed is confirmed as a useful limit that does not lead to an additional risk.