Lapping is a very complicated and random process resulting from the variation in size and shape of abrasive grains and from the numerous factors which affect the process quality. Thus it needs to be analysed by experiment rather than by a theoretical method, to obtain the relative effects of factors quantitatively. In this study, a cylindrical lapping experiment designed using Taguchi's L8 orthogonal array was performed and analysed using Yates' ANOVA table. As a result, effective factors and interaction effects were identified and discussed. Also the optimal factor combination to obtain the largest per cent improvement in surface roughness was selected and confirmatory experiments were performed.