A cohort of 101 textile mill workers was restudied with spirometry 19 months after a detailed study of their acute ventilatory response to cotton dust exposure. Information was obtained on dust exposure and smoking habits during the interval between studies. By linear regression analysis no measure of exposure to cotton dust was a significant predictor of the change in FEV(1) over the follow-up period. While the acute change in FEV(1) over the workshift in the first study appeared to be significantly associated with chronic change in FEV(1) over the follow-up interval, this association was shown to be completely explained by the effect of measurement error. The relatively short follow-up period makes estimation of the true rate of decline in FEV(1) difficult, but in this study the acute change in FEV(1) over the workshift was not useful in predicting chronic change in FEV(1).