Bipolar disorders are misdiagnosed in many cases. We hypothesized that this occurs because therapists do not base their diagnostic decision solely on criteria set out by ICD-10 or DSM-IV. We expected that instead, patients offering a plausible causal explanation for their symptoms (i.e. having fallen in love) are less likely to be diagnosed as bipolar. In the same way we expected that patients who report decreased need for sleep are more likely to be diagnosed as bipolar than those who do not present with this additional symptom. We sent a case vignette describing a person with bipolar disorder to 400 psychotherapists. This vignette was varied with respect to these two pieces of information, but each case described included all necessary criteria to diagnose a bipolar disorder according to DSM-IV or ICD-10. This variation, along with the theoretical approach of the therapist affected the likelihood of a bipolar diagnosis.