PURPOSE Rehabilitation measures have for a long time been evaluated primarily by means of pre-post comparisons of the group examined or by comparing different intervention programmes. This study was aimed at examining the current status and medium-term changes of mothers' health and psychosocial situation, comparing mothers with or without approval and participation in an inpatient mother-child rehabilitation measure. METHOD The sample consisted of 477 women. The first survey (T1) took place at the time of application; the postal follow-up survey (T2) was carried out 6 months later. At this point 353 mothers had completed their measure (IG=intervention group); 75 women had not yet started (WG=waiting group), and 49 applications had finally been rejected (AG=rejected group). This study examines how far the 3 groups differ with respect to their health parameters at the time of application as well as 6 months later. The outcome variables, containing health complaints, somatic and mental illnesses, psychosocial factors (e. g., single parenthood), or restrictions of daily activities and the range of participation in family and job affairs, were selected following the assessment guidelines of the Medical Review Board of the Statutory Health Insurance Funds (MDS). Duration and severity of symptoms were measured with Likert scales (indirect measurement of change), the mental health status with SCL-K-9. RESULTS At the time of application the 3 groups of mothers did not differ significantly with respect to their complaints and symptoms, but some minor tendencies became apparent. Mothers, who were admitted for the intervention (IG and WG) showed more complaints and symptoms than mothers, whose applications had been rejected (AG). 6 months later, highly significant changes and clinically important improvements were observed for IG mothers: large effects were achieved regarding the number of illnesses, health complaints, the degree of daily activities and participation as well as mental health of participants. Minor effects were obtained with respect to psychosocial strains. In contrast the number of illnesses and health complaints among mothers, who had not (yet) participated, increased within half a year. Similarly, psychological strains were greater than before: in the waiting group they had risen significantly, in the rejected group values resembled those of the intervention group before intervention. CONCLUSION Inpatient measures do not only cause short-term improvements of mothers' health status, they also lead to clinically relevant reductions of health impairments and complaints in the medium term. In contrast, the health status of mothers, who miss the intervention, deteriorates. This also applies to mothers, who were rejected. Therefore, it is highly recommended to take the self-reporting forms into account when identifying mothers for the intervention, and also to start the intervention early after the approval.