Teachers’ sense of selfefficacy is a hard construct to measure but is important in view of to its relationship with the quality of education. This work aims to gather evidence regarding the consistency and validity of a version of the Teachers’ Sense of Efficacy Scale (TSES) for use in Spain. The data come from a Spanish group of future earlyyears, primary and high school teachers (N = 744), and they are analysed using a multivariate approach suitable for ordinal scales, with confirmatory factor analysis (ULSMV) and decision trees (CHAID and CART). After evaluating three measurement models, evidence for the validity of a construct structure with three factors and 17 items with acceptable fit indices is provided. In addition, the convergence and consistency of the construct are both endorsed, as is the usefulness of the items for predicting overall teacher selfefficacy in the groups of undergraduate and master’s students in the sample studied.