Abstract In this article, we develop the revised and short versions of the pseudoscientific belief scale through two empirical studies ( N = 4154). This revision is motivated by the excessive length of the scale, as well as by consistent observations of poor item loadings across several studies. Exploratory factor analysis in Study 1 revealed 11 dispensable items, resulting in a 19‐item revised form, whereas in Study 2 we constructed a short eight‐item form. Confirmatory factor analysis revealed unidimensional factor structures for both scales, exhibiting excellent psychometric properties in relation to factor structure, item loadings, internal consistency and convergent validity with paranormal beliefs, conspiracy theories and need for uniqueness. Whereas the original scale provides reliable indices, we encourage the use of these improved versions to measure pseudoscientific beliefs in the context of socio‐psychological studies.