Among lead-exposed workers, there is evidence of increased mortality from chronic renal diseases (nephritis and nephrosis). Epidemiological studies using early markers of nephropathy among lead-exposed workers failed to demonstrate early renal changes. This study is aimed at assessing the glomerular function of 137 lead-exposed subjects and at evaluating whether changes in markers of glomerular function are related to exposure indices derived from longitudinal blood lead data. A control group of 153 postal workers was also investigated. Several exposure indices were derived for the exposed workers, including a time-integrated index Pb in blood (PbB) i n t and the number of times the PbB was above critical values (PbB 4 0 0 , PbB 5 0 0 , PbB 6 0 0 ). Through multiple linear regression analysis, PbB i n t was the best predictor of variation in serum β 2 -microglobulin (Sβ 2 m) and α 1 -microglobulin (Sα 1 m) and urinary albumin (UAlb). A small but statistically significant difference in the mean β 2 m was found. Sβ 2 m was also the only marker showing a significantly higher prevalence rate ratio (PRR) of abnormalities among lead-exposed workers. Though there was no clear dose-response relationship with PbB i n t as the index of dose, all the 15 subjects with abnormal Sβ 2 m in the older age group were found in the highest PbB i n t group. Furthermore, of the 8 subjects with low 4-h creatinine clearance (CrCl 4 h ), 6 had abnormal levels of β 2 m. Two subjects with CrCl 4 h of less than 75 ml/min/1.74 m 2 had high PbB i n t values, thus suggesting that high blood lead levels over a prolonged time may be associated with decreased CrCl 4 h . Though the long-term significance of elevated Sβ 2 m and UAlb is unclear, their association with high PbB i n t and decreasing CrCl 4 h indicate a potentially adverse effect. Their relationship with PbB 4 0 0 and PbB 6 0 0 suggests that the threshold of 700 μg/l for PbB may not prevent the occurrence of lead nephropathy.