To investigate the possible work-relatedness of obstructive lung disorders, 402 male Japanese employees who had continued to work in two natural mineral fiber processing factories (97% of the total number of employees) were first surveyed in 1985 and 1986. Follow-up surveys were repeated essentially annually until 1995. The surveys included forced spirometry and questionnaires concerning respiratory status. Pulmonary function in 328 workers could be examined acceptably for three years or more with at least three acceptable survey results. Correlation analyses clearly indicated that second-order height proportional values should be used to standardize conventional spirometric indices for subjects' body size difference. No consistent association was observed between pulmonary function values and room temperature at the time of measurement. Within-subject coefficient of variations of forced expiratory volume in one second and forced vital capacity were as low as 2% to 3% for data one year apart. Those coefficients of variations were almost constant throughout the entire study period. It seems critically important to minimize the measurement error of forced expiration maneuvers and keep it constant throughout all of the surveys, by striving to control the quality of pulmonary function data.