Previous investigations of workers exposed to machining fluids have shown increased rates of cough and phlegm and have shown that these exposures may cause occupational asthma. To examine acute responses to these agents, cross-shift lung function changes related to machining fluid aerosols among 89 machine operators at two factories producing automobile parts were measured and compared with the findings for 42 unexposed assembly workers studied similarly at the same factories. Workers wore a personal air-sampling device on a Monday and Friday of a working week, and spirometry was performed before and after the work shifts on both days. On Mondays, a 5% or greater decrease in the forced expiratory volume in 1-second (FEV1), regarded as an "FEV1-response," occurred in 23.6% of the machinists and in only 9.5% of the assembly workers (relative risk = 2.5, p less than .05). After adjusting statistically for a history of childhood asthma, for smoking prior to lung function testing, and for race, odds ratios for an FEV1-response of 4.4 among workers exposed to aerosols of straight mineral oils, 5.8 for oil emulsions, and 6.9 for synthetic fluids were found. The FEV1-responses on Fridays were similar to those on Mondays. There was no progressive decline in FEV1 over the work week. Personal air samples, collected with a two-stage impactor, allowed aerosol masses to be measured in three size fractions: less than 3.5 microns, 3.5-9.8 microns, and greater than 9.8 microns aerodynamic diameter. Exposure levels to each type of machining fluid were remarkably similar within each size fraction and for total aerosol levels. Total aerosol concentrations for assembly workers ranged from 0.07 to 0.44 mg/M3, and for machinists from 0.16 to 2.03 mg/m3. Inhalable particle (less than or equal to 9.8 microns) levels were derived from the sum of the air concentrations in the two smallest-size fractions, and significant cross-shift decrements in FEV1 on Mondays and Fridays were associated with inhalable aerosol levels greater than 0.20 mg/m3. These findings show that acute airflow obstruction is associated with exposures to aerosols of various machining fluids and that airway responses occur well below current recommended exposure limits.