The Raynaud's Treatment Study (RTS) exemplified clinical trials with treatments that differ qualitatively both in their modes and in their methods of delivery. The RTS compared finger-temperature biofeedback to slow-release nifedipine, a calcium channel blocker, in patients with primary Raynaud's disease. Factors influencing the study design were the nature of the interventions and control measures of the protocol, the possibility of perceived differences by the patients between the treatments once the final protocol was developed, and concern on the part of the investigators over the fact that the primary endpoint was self-reported. This paper presents the final statistical model: a double parallel design with both a placebo group and a nonspecific behavioral control group.