This work analyzes the relationship between the services clients receive in treatment programs and client ratings of program efficacy. It relies on data from a random sample of clients served by Chicago's homelessness system (N = 554). Regressions utilizing that data suggest that ratings of program efficacy are positively predicted by program ambiance, the ambiance of referral arrangements, residence in programs providing housing, and receipt of employment services. The measures of ambiance are predicted by receipt of professional services and help in locating housing. Receipt of advocacy services does not predict ratings of program efficacy nor ambiance; receipt of tangible services is negatively related to ratings of program efficacy. Results suggest that clients rate highly programs that have a positive ambiance or that provide services that clients view as immediately helpful for solving long-term needs. Clients thereby discount services known to contribute to the amelioration of homelessness. (Contains 3 tables and 1 figure.)