This paper presents the results of an experimental investigation as to the effects of grinding parameters on roundness error and surface roughness in cylindrical grinding. Many variables including the wheel materials, wheel loading and dressing, workpiece metallurgy, work drive mechanisms, work holding methods, coolant types, feeds and speeds, machine stiffness and age, surface conditions, centre conditions, floor vibrations all influence the quality of ground parts. However, the composite sum of these grinding parameters creates static and dynamic forces. It is obvious that the roundness error and surface roughness are created by many parameters, but in this study, only the effects of the depth of cut, work speed and feed rate which create the grinding forces in cylindrical grinding are investigated. The grinding experiments were planned according to the principles of orthogonal arrays (OAs), developed by Taguchi, and were performed so as to understand the effects of these parameters on roundness error and surface roughness. The experimental data was analysed by using statistical tools: the percent contribution from an analysis of variance (ANOVA) and the correlation between machining parameters with roundness error (R) and also surface roughness (Ra). Roundness was found to be the most related with the cutting speed, grinding force and depth of cut, while surface roughness is related to feed rate and work speed.