Ring crack initiation loads on glass using a spherical WC indenter were measured, and the methodology proposed by Warren and Hills and Warren was used to estimate glass toughness (K(IC)). The values obtained were overestimates of K(IC), primarily because the methodology does not account for friction correctly. Corrected results that show that the stress-intensity factor at the surface crack tip is extremely sensitive to the friction coefficient, l and to Poisson's ratio, nu of the substrate are presented. This sensitivity and an inability to obtain the minimum load for crack initiation despite numerous experimental trials, cast doubt on the utility of the technique to measure K(IC).