Twenty-one patients with chronic obstructive pulmonary disease were studied with right heart catheterization. The mean pulmonary artery pressure (??) was compared with several noninvasive tests. The pulmonary lobar diameter/thoracic ratio correlated with the ??, r = 0.677, P < 0.01. The oxygen saturation and pH were used to derive a calculated value for ??. The calculated ?? correlated with ?? values measured below 40 mm Hg, r = 0.787, P < 0.01 but correlated very poorly with ?? values measured above 40 mm Hg. The electrocardiogram revealed at least one criterion for right ventricular abnormality in all 18 patients with pulmonary hypertension, but was too nonspecific to correlate with ??. The hemoglobin concentration did not correlate with ??. There was such wide variation between the measured ?? and the various counseled. At six-month follow-up, 7% of subjects with abnormal test results had stopped smoking. Utilizing even our most cost-effective test, the SBN2, it would cost $1,392 for each "benefit" defined as one subject not smoking for six months. Application of these screening techniques is unlikely to be effective in altering smoking habits in the absence of continued physician support.