Abstract The assessment and measurement of handedness has varied across studies, limiting the comparability of results. Data from the Edinburgh High Risk for Schizophrenia Study were analyzed to investigate the effect of different methods of assessment and scoring of hand preferences on the prevalence of handedness type and on between-group differences in handedness. Handedness was measured using both the Edinburgh Handedness Inventory and the Annett Handedness Scale in 143 subjects at high risk for schizophrenia, 31 control subjects, and 27 patients with a first episode of schizophrenia. Hand preferences were identified through demonstration of items and by verbal report. No group differences were found, although the prevalence of hand preferences changed substantially depending on the definition used. Significant correlations with socio-demographic factors were found in some instances, but these correlations depended on the definition of handedness. No sex differences were identified. The magnitude of group differences remained similar, although the prevalence of handedness types varied greatly with changes in definition of handedness. Care should be taken in correlation studies to avoid spurious relationships between handedness and other factors. To allow for comparability of results across studies, researchers should adopt a standard definition of handedness.