Atypical handedness has been repeatedly reported in schizophrenia, with quantitative review of evidence showing an increase of non-right-handedness in patients diagnosed with schizophrenia. Mixed-handedness is also higher among non-clinical people scoring high on questionnaires aimed at measuring psychosis-proneness. However, the greatest part of information on non-clinical samples came from samples collected in North America or in the UK: differences by countries in the socio-cultural pressure to use the right hand could influence the results. In this study 604 Italian non-clinical participants (248 males, 41.1%; 356 females, 58.9%; mean age=34.5+/-11.9) completed the Annett Hand Preference Questionnaire (HPQ), the General Health Questionnaire (GHQ), and the Peters et al. Delusions Inventory (PDI). In the sample, 527 subjects (87.3%) were classified on the HPQ as right-handed, 53 (8.8%) were classified as mixed-handed, and 24 (4.0%) were classified as left-handed. Mixed-handed scored statistically higher on the PDI than the right-handed and left-handed, but right-handed and left-handed did not differ from each other on a statistical ground. The difference by handedness was specific for PDI, since scores on the GHQ-12 did not differ by handedness group. The links between mixed-handedness and psychosis-proneness in non-clinical samples are a reliable finding, deserving further investigation as a model for the risk of schizophrenia.