This paper examines the nature and types of information and syntheses relevant to the strategic and tactical decisions required in the treatment of depression with medication. Strategic decisions (What to do?) include determining what is wrong, defining treatment objectives, selecting the first treatment, deciding the next best treatment (should the first treatment fail), and deciding when to discontinue a successful treatment. Tactical decisions (How to do?) include determining where and by whom treatment is conducted, and the specific steps involved (e.g., dosing, frequency of visits, assessment of outcome, recognition and management of side effects, compliance, etc.) in the treatment as well as in the discontinuation of treatment (e.g., taper schedule, frequency of visits, etc.). The treatment of each patient is inherently an experimental single case study. Because individual patients differ, the most luxuriant data bases cannot fully address all strategic and tactical decisions with certainty. Thus, practitioners must apply general principles and group (nomothetic) data to individual patients. The relationship between research efforts and the information needed by clinicians is examined. Suggestions for developing and reporting more clinically relevant information from efficacy trials and other research are made.