The verdict of not guilty by reason of insanity (NGRI) continues to be the subject of much debate (cf. Perr, 1985; Steadman & Morrisey, 1986). In theory, insanity acquittees are sick and in need of treatment. They are not to be punished because they were not responsible for their offenses. But in practice, most insanity acquittees are involuntarily detained for indefinite periods of time in secure hospitals (Pasewark, 1986; Pogrebin, Regoli, & Perry, 1986). What factors are related to length of detention for these men? How do their lengths of detention compare to those of persons found guilty of similar offenses? Cooke and Sikorski (1974) found several factors to be related to length of institutionalization, including marital status, educational and occupational status, race, offense, diagnosis, previous offense history, and previous history of hospitalization. However, this study has been criticized for using excessively liberal criteria for statistical significance (Pasewark, 1986). In another study (Steadman, Pasewark, Hawkins, Kiser, & Bieber, 1983), multiple regression techniques were used to predict length of hospitalization for a group of insanity acquittees in New York State. It was found that nine variables (the most important being severity of offense, marital status, and sex) were related to length of stay, but that the nine variables accounted for only 11% of the variance in the length of stay. Neither of the above studies included treatment variables. Steadman et al. (1983) concluded that variables related to therapeutic progress might be the most important ones for predicting length of stay. There have been several other studies in which the focus has been a comparison between insanity acquittees and persons convicted of similar acts. Results regarding length of stay for the two groups have been inconsistent, with some studies finding that insanity acquittees stayed shorter lengths of time in custody than their convicted counterparts (Pasewark, Pantle, & Steadman, 1982; Phillips & Pasewark, 1980), some finding the two groups stayed equivalent lengths of time (Braff, Arvanties, & Steadman, 1983; Pantle, Pasewark, & Steadman,