Purpose: The purpose of this paper is to analyze the construct validity of learning transfer system inventory (LTSI) for use in Portugal. Furthermore, it also aims to analyze whether LTSI dimensions differ across individual variables such as gender, age, educational level and job tenure. Design/methodology/approach: After a rigorous translation process that included forward and backward translation, subjective evaluations of the translated items, and pilot testing, the Portuguese version of LTSI was completed by 484 trainees, who attended different kinds of training courses. Two separate exploratory factor analyses were run to analyze the specific and general domains covered by LTSI. An initial analysis with the validated items and a second one with the validated and research items were also completed. Findings: The results of the initial factor analyses showed a 16-factor structure that accurately paralleled the 16-factor structure of the original English version and revealed that this one is the most parsimonious factor structure. Statistically significant differences in some LTSI dimensions were found across gender and education-level variables. Research limitations/implications: Data were collected through self-reported measures using a single instrument applied immediately after the training, which might have affected the internal validity of the study. Although this study covered a wide variety of trainees' business areas and types of training, it might be possible that some types of organizations and training were not included in the analysis, thus limiting the generalization of the results to those contexts. Practical implications: Based on the results of this study, Portuguese companies are now able to develop their training evaluation practices and to assess the training transfer factors through a valid and reliable instrument. Originality/value: The paper expands LTSI validity in Europe, demonstrating that it has construct validity for use in Portugal. (Contains 10 tables.)