The method of Guo et aL (AnaL Chim. Acta, 1997, 349, 313-318) for the determination of the toxicologically relevant arsenic in urine was verified and then used for the determination of arsenic in urine of the Czech population for monitoring purposes. Statistical evaluation at the level alpha = 0.05 did not prove any significant differences between industrial and agricultural regions, between males and females and smokers and nonsmokers. Likewise no differences were found among children in all the regions monitored. In the adult population small differences were found between some regions but these differences were not dependent on industrial pollution. The values of toxicologically relevant arsenic are low for all regions. The summarised value of the median for all groups together is 3.5 microg (g creatinine)(-1).