This study examined variables that predicts case managers' assessment of th
e strength of the helping alliance with people with a psychiatric disabilit
y one year after the initiation of treatment. Although some background char
acteristics and treatment process variables were significant predictors of
the helping alliance in bivariate analyses, these variables were not signif
icant predictors of the helping alliance in multivariate analyses. However,
changes in self-reported outcomes (e. g., homelessness, alienation, and in
terpersonal adjustment) were significantly related to the case managers' as
sessment of the strength of the helping alliance. Thus, it appears that the
causal relationship between the strength of the helping alliance and impro
vement of the person with a psychiatric disability may be largely reciproca
l.