A total of 645 cotton mill operatives were administered a respiratory questionnaire. Of these, 85 (13.2%) complained of one or more work-related respiratory symptoms: 23 (3.6%) had byssinosis and the remaining 62 had symptoms not conforming to byssinosis (nonbyssinotic symptomatics, NBS). All byssinotic, 56 NBS, and 84 matched asymptomatic operatives underwent pulmonary function testing (FEV1 and FVC), skin testing to common allergens, and histamine bronchial challenge. Work area and personal breathing zone cotton dust concentrations were assessed, and a cumulative cotton dust exposure index was calculated for each individual. Byssinotic, NBS, and asymptomatic operatives all had reduced FEV1; observed mean liters (95% CI); predicted mean: byssinosis, 2.36 (2.09 to 2.63), 3.02; NBS, 2.94 (2.71 to 3.17), 3.29; and asymptomatic, 3.12 (2.95 to 3.29), 3.31. Only byssinotic subjects had evidence of impaired FVC: 3.31 (2.97 to 3.65), 3.69. The majority of byssinotic operatives (18 of 23) had bronchial hyperreactivity (BHR) in comparison with 21 of 56 NBS and 14 of 84 asymptomatic operatives. Mean log PD20 (95% CI) values were significantly lower in the byssinotic group -0.72 (-1.42, -0.02) than in NBS 0.57 (0.08, 1.06) and asymptomatic subjects 0.57 (-0.26, 1.39). The distribution of atopy did not differ significantly between groups, and lung function did not differ significantly between atopic and nonatopic subjects. The cumulative cotton dust exposure index was the only dust parameter to be significantly greater in those with BHR (mean mg-yr/m3 [95% CI] 14.13 [13.1 to 15.1]) than those with normal reactivity [5.35 (3.9 to 6.8)].