Studies comparing voluntary and involuntary psychiatric patients were quantitatively reviewed in order to identify antecedents and consequences of civil commitment. Data for the review consisted of correlations between commitment status and 15 patient variables, including demographic, social, behavioral, psychiatric, and hospitalization characteristics. The correlations were examined separately for studies from different time periods (1950s-1960s vs. 1970s-1980s) and different research settings (state hospital vs. other). The correlates of commitment appeared to vary as a function of these factors. However, all of the older studies were conducted in state hospitals, whereas the majority of more recent studies were carried out in non-state hospital settings. Because of this confounding, it is unclear whether the correlates of commitment have changed over time or simply differ from one setting to another. Limitations of the existing research and directions for future inquiry are discussed. (57 refs)