• The effectiveness of in vivo exposure is evaluated in a comprehensive review of outcome studies on phobia treatments. A strictly data-based analysis focuses on issues of analog vs clinical research, methodological quality of designs, utilization of generalization measures, and maintenance of therapeutic change. In vivo exposure appears to be an effective treatment, both short- and long-term, and some decision rules for the clinical application are provided. However, from a researcher's point of view many methodological weaknesses (such as frequent lack of follow-up; unrealistic behavioral measures; and the utilization of unnecessarily complicated, and at times confusing, designs) are noted. A considerable consensus on the major procedural variables and on positive overall outcome justifies the clinical utility of the procedure. Detailed suggestions for further treatment comparisons are outlined, and a cautionary note on the potential of such a data-based review are added.