In a previous study of new miners from the National Study of Coal Workers' Pneumoconiosis (NSCWP), researchers examined changes in spirometry values associated with coal mine dust exposure (Br J Ind Med 1993; 50:929-937). An unusual pattern of dust-related effects was observed: initial sharp decrements in FVC and FEV1 were followed by partial recovery. In the current study, similar methods were used to analyze data from experienced miners. Each of 1,915 male subjects contributed data from two of the NSCWP field surveys: either Round 1 (1969-71) and Round 2 (1972-75) and Round 4 (1985-88). From the cross-sectional analysis at Round 1 or Round 2 (R1/R2), changes of +0.6 ml FVC and -0.5 ml FEV1 were associated with each mg/m3-yr of cumulative coal mine dust exposure, but were not statistically significant (p > 0.05). From the analysis of longitudinal change in spirometry from R1/R2 to Round 4 (R4), annual declines in FVC (-0.10 ml/yr per mg/m3-yr, p = 0.003) and FEV1 (-0.07 ml/yr per mg/m3-yr, p = 0.006) were associated with pre-R1/R2 exposure. Both the pattern and the magnitude of the exposure-response relationship were different for experienced versus new miners. Possible reasons for these contrasts include differences in cumulative exposure between the two groups and the healthy worker effect among experienced miners.