Antidepressant medication has apparently become the most popular treatment for depression in the USA. Several beliefs about the efficacy of antidepressant medications prevail among mental health professionals and the public. This paper explores relevant research data and raises questions about these beliefs. Many of the common beliefs about these medications are not adequately supported by scientific data. The following issues are raised: (1) industry-funded research studies which result in negative findings sometimes do not get published; (2) placebo washout procedures may bias results in some studies; (3) there are serious questions about the integrity of the double-blind procedure; (4) the 'true' antidepressant drug effect in adults appears to be relatively small; (5) there is minimal evidence of antidepressant efficacy in children; (6) side effects are fairly common even with the newer antidepressants; (7) combining medications raises the risk for more serious complications; (8) all antidepressants can cause withdrawal symptoms; (9) genetic influences on unipolar depression appear to be weaker than environmental influences; (10) biochemical theories of depression are as yet unproven; (11) biological markers specific for depression have been elusive; (12) dosage and plasma levels of antidepressants have been minimally related to treatment outcome; (13) preliminary evidence suggests that patients who improve with cognitive-behavioral psychotherapy show similar biological changes as those who respond to medication, and (14) the evidence suggests that psychological interventions are at least as effective as pharmacotherapy in treating depression, even if severe, especially when patient-rated measures are used and long-term follow-up is considered.