According to Meehl's model of schizotypy, there is a latent personality organization associated with the diathesis for schizophrenia that can be identified in several ways. We sought to examine the structural invariance of four Chapman psychosis-proneness scales (CPPS) across three groups of putative schizotypes, namely, clinically-, biologically-, and psychometrically-identified schizotypes. We examined the factor structure of the Perceptual Aberration (PER), Magical Ideation (MIS), Revised Social Anhedonia (RSAS), and Revised Physical Anhedonia (RPAS) scales in 196 schizophrenia patients, 197 non-psychotic first-degree relatives, and 1,724 non-clinical young adults. The confirmatory factor analyses indicated that the best-fitting model was one in which there is a two-factor model with negative schizotypy (RSAS and RPAS) and positive schizotypy (PER and MIS). All three samples fit the model well, with Comparative Fit Indices>0.95 and Tucker Lewis Indices>0.90. The root mean square error of approximations were all small (P values⩽0.01). We also observed that for both anhedonia scales, the groups' mean scale scores varied in the hypothesized direction, as predicted by Meehl's model of schizotypy. All three Chinese samples, namely, the patients (clinical schizotypes), relatives (biologically-identified schizotypes), and non-clinical young adults (containing psychometrically-identified schizotypes) showed the same factorial structure. This finding supports the suitability of the CPPS for cross-cultural and/or genetic investigations of schizotypy.