A review of 47 clinical trails of Prozac and three other anti-depressants has suggested that they are not helping most of the patients for whom they are prescribed. The unpublished trails particularly suggest that the benefits of anti-depressants, in the words of the researchers, fall 'below the accepted criteria for clinical significance’.
Speaking as someone who has taken these drugs, I'm not sure. The real problem is not the drugs but the way they are promoted and prescribed. In the right circumstances anti-depressants can do a job. The trouble is that we're doling them out like sweets - 31 million prescriptions in 2006 alone.