While an increased awareness of the dehumanizing and organizationally self-defeating qualities of residential institutions has fostered a movement toward community-based care, little research has been done to measure the "institutionality" of the alternate settings. As a part of a larger study aimed at describing alternative settings for juveniles, two dimensions of institutionalization, i.e., perceived organizational control and institutional program design, were examined in 30 settings in six different states. Perceptions of organizational control were collected from 134 members (72 percent line staff, 28 percent supervisory personnel) across all sites, using a matrix charting the level of influence in treatment decision making for all staff members. Four clusters of sites differing on perceived control were identified ranging from an administrative hierarchy (N=6), to a middle-level control model (N=3), to a resident/supervisor model (N=9). Sites were also clustered according to 19 items representing variables related to program rules. A cross-tabulation of each site's membership in the two clustering solutions (the organizational control measure and the program variables) showed that case frequencies were distributed nearly randomly; congruence between the cluster solutions appeared nearly nonexistent. That the two measures were unrelated may reflect measurement inadequacy, or may mean that the two constructs are only marginally related to each other. (MCF)