After a report in 1980 of the first three diagnosed locally cases and a preliminary epidemiological investigation that found little evidence of the disease, a survey was aimed at determining the prevalence of byssinosis in Hong Kong. Some 1776 workers in six cotton mills were studied using the standard MRC questionnaire and portable spirometers. Only 48 (2.7%) of the mill workers had symptoms acceptable for a diagnosis of byssinosis. The pattern of relation to dust exposure levels was similar to findings in other countries: blowing and carding process operatives had twice the prevalence rate of the spinners. Another 178 workers (10%) had symptoms of chest tightness or breathlessness or both that were not related to the first exposure after a break and therefore did not fit the standard diagnosis. Some 257 workers (14.5%) had chronic obstructive airflow disease but only 12 (4.7%) had chronic bronchitis. Job mobility had self selection of sensitive cases out of cotton dust exposure seem the most likely explanations for the low prevalence. The significance of non-specific lung ailments needs further assessment to elucidate the possible connection with cotton dust exposure.