To determine if pulmonary vessels develop vasospasm during Raynaud's phenomenon, digital vasospasm was induced by hand immersion in 15 degrees C water (cold pressor test) in 17 subjects, and pulmonary function was measured during the subsequent 120 minutes. Five healthy persons were control subjects, seven subjects had well documented systemic disorders associated with Raynaud's phenomenon (secondary Raynaud's), and five subjects had a history of Raynaud's phenomenon but no evidence of an associated disorder (primary Raynaud's). The only measure of pulmonary function that changed significantly following cold pressor testing was carbon monoxide diffusing capacity. Subjects with primary Raynaud's phenomenon had normal baseline carbon monoxide diffusing capacity (23.7 +/- 4.6 ml/minute/mm Hg) but demonstrated significant decreases (p less than 0.05) at 15 minutes (21.2 +/- 3.5 ml/minute/mm Hg), 45 minutes (19.5 +/- 3.7 ml/minute/mm Hg), and 120 minutes (17.1 +/- 2.1 ml/minute/mm Hg) after cold pressor testing. Subjects with secondary Raynaud's phenomenon had low baseline carbon monoxide diffusing capacity (71 percent predicted) and showed no significant change following cold pressor testing. These findings indicate that digital vasospasm in patients with primary Raynaud's phenomenon is part of a systemic vascular response that includes a decrease in the size of the pulmonary capillary bed.