4,4'-Methylenebis-(2-chloroaniline) (MOCA) is used in the manufacture of polyurethane. The IARC classifies MOCA as a probable human carcinogen. Suggested changes to guidelines for health surveillance of MOCA-exposed workers in Australia include a reduction in acceptable levels of urinary MOCA to below 15 mumol/mol creatinine. Twelve male workers aged 24 and 42 years were recruited into this study from four work locations where MOCA is used. Exfoliated urothelial cells from prework urine samples on a midweek work day were assessed for micronucleus (MN) frequencies. Postwork urine samples were analysed for total MOCA. Blood samples collected on the same day were cultured for 96 h and cytochalasin-B-blocked cells were scored for MN. Eighteen male control subjects (23-59 years) provided corresponding urine and blood samples. Median urinary MOCA concentrations were 6.5 mumol/mol creatinine (range 0.4-48.6 mumol/mol creatinine) in postwork samples of MOCA-exposed workers. MOCA was not detected in urine of control workers. Mean MN frequencies were higher in urothelial cells and lymphocytes of MOCA workers (14.27 +/- 0.56 and 13.25 +/- 0.48 MN/1000 cells) than in controls (6.90 +/- 0.18 and 9.24 +/- 0.29 MN/1000 cells). The mean number of micronucleate cells was also higher in both tissues of exposed workers (9.69 +/- 0.32 and 8.54 +/- 0.14 MN cells/1000 cells) than in controls (5.18 +/- 0.11 and 5.93 +/- 0.13 MN cells/1000). There was no correlation between postwork urinary MOCA concentrations and MN frequencies in either tissue. This study suggests that exposures to MOCA in South Australia are similar to those of a decade ago and are at levels similar to those currently acceptable in Australia. These are associated with genotoxic effects in urothelial cells and peripheral blood lymphocytes. It may be prudent to reduce MOCA exposures in line with proposed guidance values.