Evaluation of EM methods for measuring airborne asbestos fiber concentrations and size distributions was carried out by studying a large number of variables and subprocedures in a five-phase program using elaborate statistically designed experiments. Observations were analyzed by advanced regression techniques to evaluate the effects of independent variables and subprocedures. It was shown that the optimized method for estimating airborne chrysotile should have the following features: (a) collecting an air sample on Nuclepore filter; (b) coating the Nuclepore filter with carbon; (c) transferring the particulate deposit to a 200-mesh electron microscope grid using chloroform in a modified Jaffe-wick washer; (d) examining the grid at about 10,000 x magnification (20,000 x for counting very fine fibers); (e) counting fibers using a field of view method; and (f) identifying the type of asbestos from morphology and selected area electron diffraction. A provisional manual of instructions was prepared (PB-276 481) and six independent laboratories participated in an interlaboratory test of the proposed method using two air samples. One of these was prepared at IITRI from pure aerosolized UICC chrysotile, and the other was an ambient air sample collected by IITRI personnel in a factory processing asbestos. Intercomparison of the results from the separate laboratories yielded some preliminary estimates of the precision and accuracy of the provisional method.