The aim of this study was to examine whether the ROCF qualitative scoring system developed by Loring, Lee and Meador for differentiating complex partial seizures originating from either the right or left temporal lobe is effective in differentiating left-sided and right-sided brain lesions that are the result of cerebrovascular insult or brain tumours. We were also interested in determining whether this scoring system, which was developed for scoring trials with delayed recall, could be applied to trials with immediate recall and copying. The study consisted of 24 participants with lesions of the left hemisphere and 33 participants with right-sided lesions. Participants with right-sided lesions had a significantly greater number of qualitative errors in copying, immediate and delayed recall, and these three variables are the major contributors in distinguishing between groups. Based on these variables and quantitative results on copying, immediate and delayed recall, we were able to correctly classify 78.3 % of participants with left-sided lesions and 66.7 % of participants with right-sided lesions. Given that more than 90 % of participants with left-sided lesions had two or more errors in delayed recall, it is clear that the criterion of two or more errors which was set by Loring, Lee and Meador for patients with a right-sided focus in epilepsy is not applicable to patients with tumours and strokes. These results do not confirm the usefulness of qualitative errors for distinguishing left-sided and right- sided lesions caused by a tumour or stroke, except, perhaps, in cases of very high results (six or more errors) and when one of these errors is error X in immediate and/ or delayed recall.