OBJECTIVE The aim of the present study is to investigate whether lead (Pb) in urine (Pb-U) can be a valid surrogate of lead in blood (Pb-B), the traditional biomarker of exposure to lead in occupational health. METHODS Blood and spot urine samples were collected from 258 workers of both sexes occupationally exposed to lead. The samples were analyzed for lead by graphite furnace atomic absorption spectrometry, and the correlation between Pb-B and Pb-U was examined by linear regression analysis before and after logarithmic conversion. RESULTS The correlation coefficient (0.824; P < 0.01) was largest when the relationship between Pb-B and Pb-U was examined with 214 cases of one sex (i.e., men) after Pb-U was corrected for a specific gravity (1.016) of urine (Pb-Usg) and both Pb-B and Pb-Usg were converted to logarithms. The geometric means (GMs) of Pb-B and Pb-Usg for the 214 men were 489 microG/l and 81 microg/l, respectively. When Pb-Usg was assumed to be 100 microg/l in this set of correlations, the 95% confidence range of Pb-B for the group mean was narrow, i.e., 543-575 microg/l (with GM of 559 microg/l), whereas that for individual Pb-B values was as wide as 355-881 microg/l. CONCLUSIONS The correlation of Pb-U with Pb-B among workers occupationally exposed to Pb was close enough to suggest that Pb-U may be a good alternative to Pb-B on a group basis, but not close enough to allow Pb-U to predict Pb-B on an individual basis.