Previous results have shown that when asbestos fibres on membrane filters are counted over the whole field of view of the microscope, lower concentration values are obtained than when the counting is confined to a central graticule grid area. Further studies have been carried out to verify this grid effect, to determine its magnitude for amosite and chrysotile asbestos fibres, and to ascertain its causes. The effects of various physical and human factors were investigated.With small graticule grids, fibre counts were increased by a factor of about 1.5 for amosite and 2.5 for chrysotile as compared with full-field counting. The effect was attributed mainly to human errors associated with full-field counting. It was concluded that more reliable and accurate counts are obtained in the routine evaluation of airborne asbestos dust concentrations by using an eyepiece graticule method of counting than by the full field method. It is therefore recommended that this method of counting be adopted and that steps be taken to reach national or, preferably, international agreement on a standard grid counting procedure.