OBJECTIVES: Treatment Satisfaction (SATMED-Q) is a generic PRO instrument measuring Treatment satisfaction with medicines in chronic health conditions. Several multisample analysis are carried out in order to assess the stability of the questionnaire structure across of different types of patients, and to discover possible differences in dimension correlations for different drug treated pathologies. METHODS: Four samples of patients with different known pathologies following regular medication treatment were gathered from specialized units, and primary attention centres: hypertensive, drug-refractory-epilepsy subjects, caregivers of patients with dementia of the Alzheimer type and a normative sample. Exploratory factor analyses on each separate sample were carried out to assess sample specific structures. Multi-sample confirmatory factor analysis was used to assess the degree of agreement between sample structures using the known theoretical structure. Congruence statistics were also used with exploratory structures. RESULTS: A total of 2777 patients were gathered: 1,025 hypertensive patients -47% women, 63 (±11.3) years old-, 768 epileptic patients -65% women, 41 (±13.4) years old-, 842 Alzheimer patients caregivers -63% women, 78 (±6,8) years old-, and the normative sample used to develop the SATMED-Q composed by 442 patients of different pathologies -50% women, 62 (±13.6) years old-. The theoretical 6-factor solution identified in the original instrument was attained in all sample specific exploratory analyses. The multi-sample confirmatory analysis matched the 6-dimensional solution but usual goodness of fit problems with large sample sizes were found. Sample specific confidence intervals for factor loadings and factor correlations were obtained. Congruence statistics were reasonable, but were not able to explain differences in the obliqueness of factors. Meaningful differences in dimension correlations between samples were found. CONCLUSIONS: The structure of the SATMED-Q questionnaire is shown to be stable across different type of treatments and patients, but the relations between dimensions vary depending on the pathology studied.