While much research has suggested that the integrity of the blind is compromised in psychotropic drug trials, little research has been conducted on blinding in substance abuse trials. The current study examines the integrity of the blind in an outpatient pharmacotherapy trial investigating the effectiveness of amantadine and propranolol in treating cocaine addiction. Results suggest that neither nurses (N = 174, kappa = 0.08, p = 0.22) nor participants (N = 163, kappa = 0.09, p = 0.26) could accurately predict treatment assignment. Furthermore, nurses' perceptions of treatment assignment were significantly related to trial completion, medication compliance, and cocaine use--results that may have training implications for medical personnel.