Abstract Background Patient empowerment is associated with improvements in different patient-reported and clinical outcomes. However, despite being widely researched, high quality and theoretically substantiated disease-generic measures of patient empowerment are lacking. The few good instruments that are available have not reported important psychometric properties, including measurement invariance. The aim of this study was to assess the psychometric properties of the 15-item Gothenburg Empowerment Scale (GES), with a particular focus on measurement invariance of the GES across individuals from three countries. Methods Adults with congenital heart disease from Belgium, Norway and South Korea completed the GES and other patient-reported outcomes as part of an international, cross-sectional, descriptive study called APPROACH-IS II. The scale’s content (missing data) and factorial validity (confirmatory factor analyses), measurement invariance (multi-group confirmatory factor analyses), responsiveness (floor and ceiling effects) and reliability (internal consistency) were assessed. Results Content validity, responsiveness and reliability were confirmed. Nonetheless, metric but not scalar measurement invariance was supported when including the three countries, possibly because the scale performed differently in the sample from South Korea. A second set of analyses supported partial scalar invariance for a sample that was limited to Norway and Belgium. Conclusion Our study offers preliminary evidence that GES is a valid and reliable measure of patient empowerment in adults with congenital heart disease. However, cross-country comparisons must be made with caution, given the scale did not perform equivalently across the three countries.