In epidemiological studies designed to identify potential health risks of exposures to synthetic vitreous fibers, the characterization of airborne fiber dimensions may be essential for assessing mechanisms of fiber toxicity. Toward this end, air sampling was conducted as part of an industry-wide study of workers potentially exposed to airborne fibrous dusts during the manufacture of refractory ceramic fibers (RCF) and RCF products. Analyses of a subset of samples obtained on the sample filter as well as on the conductive sampling cowl were performed using both scanning electron microscopy (SEM) and transmission electron microscopy (TEM) to characterize dimensions of airborne fibers. Comparison was made of bivariate fiber size distributions (length and diameter) from air samples analyzed by SEM and by TEM techniques. Results of the analyses indicate that RCF size distributions include fibers small enough in diameter (< 0.25 microm) to be unresolved by SEM. However, longer fibers (> 60 microm) may go undetected by TEM, as evidenced by the proportion of fibers in this category for TEM and SEM analyses (1% and 5%, respectively). Limitations of the microscopic techniques and differences in fiber-sizing rules for each method are believed to have contributed to the variation among fiber-sizing results. It was concluded from these data that further attempts to characterize RCF exposure in manufacturing and related operations should include analysis by TEM and SEM, since the smallest diameter fibers are not resolved with SEM and the fibers of longer length are not sized by TEM.