This study aims at calibrating the Social Performance Schedule Scale Revised (SPSS-R) in a sample of 181 adults with intellectual disability using the Rasch-Andrich Rating Scale Model with the purpose of elucidating its structure, dimensionality, and psychometric properties. In order to obtain a version with a clear factorial structure and coherent with the analysed construct (social skills), it was necessary to proceed with a drastic reduction of the scale. The resulting scale, composed of 24 items, showed the emergence of two distinct dimensions. Its psychometric properties (reliability and validity) turned out to be adequate, and it showed correct items and individuals fit to the Rasch model. According to their difficulty, the items concentrated in the mean zone of the analysed dimensions, which makes it advisable to include, in future studies, items in the high and low range of the latent variable.