The aim of the present study was to optimize the procedures for urinary mutagenicity testing in order to lower the baseline variation in mutagenic activity found in urine from unexposed subjects and to increase the sensitivity of the method. This was accomplished by using urine from nonsmokers and smokers as well as chemically spiked nonsmokers' urine. Diet was standardized. The number of mutants per ml of urine calculated from the linear portion of the dose-response curve was used as a measure of mutagenicity. The parameters investigated were (i) the total volume of urine per resin volume, (ii) the flow rate, (iii) the pH, (iv) the ionic strength of the urine, and (v) elimination of histidine. XAD-2 and C18 Sep-Pak resins recovered mutagens in smokers' urine and in chemically spiked urine with the same efficiency when an optimized procedure was adopted. The optimized procedure using a maximum volume of 50 ml acidified urine per Sep-Pak cartridge, or equal amount of XAD-2 resin, gave well over ten times greater recovery of mutagens from smokers' urine than in earlier reports. Histidine was effectively eliminated, and the background variation was also lowered.