VESPAR is a test of reasoning designed to measure fluid intelligence in neurological patients with more accuracy than has hitherto been possible. It comprises six sections, three matched sets of verbal and spatial reasoning problems, each dedicated to one of three forms of inductive reasoning: odd one out, by analogy and series completion. The VESPAR overcomes the constraints of classical tests of reasoning by using simple stimuli that are more readily accessed by patients who have suffered physical or cognitive impairments as a result of neurological illness. It does so by not timing performance, by using high frequency stimulus words and by using visually distinct spatial stimuli. A multiple choice format has been adopted to reduce both short term memory load and output demands on the patient. The reasoning items are all novel, thus the contribution of stored knowledge and procedures are reduced to a minimum. The assessment of patients for neurodiagnostic and neurorehabilitation needs will be facilitated. Patients with motor disabilities or speech production impairments will be less handicapped than on other tests, by the VESPAR's requirement for simple "pointing" response. A measure of fluid intelligence can be obtained in patients with aphasia on the spatial section of the test and in patients with visual and spatial problems on the verbal section. The VESPAR's low reliance on learnt knowledge makes it a good indicator of innate ability, relatively unaffected by educational experience. Thus, although originally developed for adult neurological populations, the test is suitable for a wide range of clinical, educational, occupational and research applications.