The purpose of this study was to systematically review non-randomized controlled trials of balneotherapy effects on locomotorium diseases in order to assess their quality and organize the evidence.
 We searched the databases of PubMed, CINAHL, Web of Science, JDream ll, and lchushi-Web forpapers published from 1990 to December 2, 2008. Eligible studies were non-randomized controlled trials in which balneotherapy exclusive of underwater exercise was the primary intervention. There was no limitation on the language in which the paper was written.
 We found only 4 papers (2 in English, 1 in Croatian, and 1 in Japanese) that fulfilled the eligibility criteria. These papers differed in the type of diseases studied, so a meta-analysis could not be applied. Their individual results showed that hot springs treatment in combination with a comprehensive fitness class was more effective than hot springs treatment alone, and that balneotherapy had therapeutic effects on arthrosis deformans, psoriatic arthritis, and lumbago. However, it was difficult to accurately interpret the effects because these papers were deficient in providing descriptions of important details of the studies.
 To improve the quality of future investigations performed with non-randomized controlled trials, we suggest that investigators design a study based on various checklists, implement interventions, conduct evaluations, finally write papers summarizing their findings.