Determinations of exposure-response relationships between crocidolite and the major asbestos-related diseases in the Wittenoom cohort have previously depended on the validity of estimates of airborne exposure to asbestos. This work aims to validate the airborne exposure measurements by obtaining measurements of the concentrations of uncoated crocidolite fibers and asbestos bodies retained in the lungs of individual workers, and to estimate the half-life of crocidolite fibers in the lungs. Samples of lung tissue, excluding tumor, of all former Wittenoom workers known to have died in Western Australia (WA) were sought from teaching hospitals, pathology departments, and the Coroner's pathologist. The lung specimens were processed using Pooley's method with TEM for counts of fibers of all types and using Smith and Naylor's method with conventional light microscopy for asbestos bodies (AB). Multiple linear regression was utilized to examine the associations between crocidolite concentrations in the lung and duration of employment at Wittenoom, time since last employed at Wittenoom, nature of job, estimated average fiber concentration at the worksite, and estimated cumulative crocidolite exposure (CCE) in fiber-years/ml for each subject. Lung tissue from 90 cases was processed and there was good agreement between counts of crocidolite fibers, asbestos bodies, and CCE. Correlations were 0.77 for AB and fibers, 0.54 for AB and CCE, and 0.58 for CCE and fibers, after log transformation. The half-life of crocidolite fibers in the lung was estimated at 92 months (95% CI 55-277 months). Previous estimates of airborne exposure to Wittenoom crocidolite have been reasonably reliable. The relatively simple technique of light microscopy for counting ABs in lung tissue also provides a useful and reliable indication of the level of past occupational exposure to crocidolite in subjects whose exposure has been only to crocidolite. The half-life of crocidolite fibers in the lungs of former Wittenoom workers is about 7-8 years.