We assessed blind integrity in a double-blinded study comparing paroxetine 20 mg with inert placebo in 20 volunteer subjects who were attempting to stop using methamphetamines. At the end of the study, the blinded clinicians reviewed subject charts and attempted to identify the assigned conditions for the 13 subjects who completed two or more weeks of the study. The three subjects who completed the entire study also attempted to identify their conditions on a questionnaire. We conclude that the blind may unwittingly be broken when the treatments under study are placebo and the selective serotonin reuptake inhibitor (SSRI) paroxetine. The integrity of the blind should be tested in all double-blind SSRI studies.