Pre- and postshift spirometry was obtained on 1,113 blue collar workers employed at 35 work sites judged to have no hazardous occupational respiratory exposures on the basis of inspection visits and environmental sampling. In addition to spirometry, a standardized questionnaire was administered by trained personnel. A study population of 944 remained after exclusion of workers for incomplete demographic data and/or spirometry with poor within-session reproducibility, i.e., greater than or equal to 10% variability in the two largest values of either FVC and/or FEV1. Overall mean values of changes across the work shift in FEV1 and peak expiratory flow rate (PEFR) were -0.8% (-0.04 L) and +2.1% (+0.13 L/s), respectively. Standard deviations for these across-shift changes were 5.8% (0.19 L) and 13.2% (1.19 L/s) for FEV1 and PEFR, respectively. In univariate analyses, mean values of across-shift changes were not statistically related to age, race, sex, smoking status, work shift, or FEV1/FVC ratio. However, variability (i.e., standard deviation) of across-shift changes were significantly related to some of these factors. These observations provide a basis for interpreting results of occupational respiratory morbidity surveys involving measurement of changes in FEV1 and/or PEFR across a work shift.