This investigation provides information concerning the risk or respiratory malignancy in relation to duration, degree, and fiber type of exposure to asbestos in a manufacturing cohort of 5,645 with long-term follow-up. Excess mortality for this cause was found in groups with moderate and high cumulative exposure (standard mortality ratios of 290 and 226). Analysis of the influence of components of total exposure dose (duration, average concentration) revealed no detectable excess risk in persons employed for less than 2 years or with low-degree average exposure. Exposure to corcidolite (blue) fiber in addition to the predominantly used chrysotile in pipe making appeared to be associated with higher risk than was exposure to chrysotile alone.