AIMS: To determine the association between lung function of coke oven workers and exposure to coke oven emissions. METHODS: Lung function data and detailed work histories for workers in recovery coke ovens of a steelworks were extracted from a lung function surveillance system. Multiple regressions were employed to determine significant predictors for lung function indices. The first sets of lung function tests for 613 new starters were pooled to assess the selection bias. The last sets of lung function tests for 834 subjects with one or more year of coke oven history were pooled to assess determinants of lung function. RESULTS: Selection bias associated with the recruitment process was not observed among the exposure groups. For subjects with a history of one or more years of coke oven work, each year of working in the most exposed "operation" position was associated with reductions in FEV(1) of around 9 ml (p = 0.006, 95% CI: 3 ml to 16 ml) and in FVC of around 12 ml (p = 0.002, 95% CI: 4 ml to 19 ml). Negative effects of smoking on lung function were also observed. CONCLUSIONS: Exposure to coke oven emissions was found to be associated with lower FEV(1) and FVC. Effects of work exposure on lung function are similar to those found in other studies.