BACKGROUND: Pulmonary artery waveforms fluctuate because of changes in intrathoracic pressure caused by respirations. Monitoring system algorithms determine digital displays of pressure measurements on the basis of recognition, analysis, and comparison of consecutive waveforms. OBJECTIVE: To compare three methods of measuring pulmonary artery pressure during mechanical ventilation and spontaneous breathing in cardiac surgery patients with stable hemodynamics. METHODS: Pulmonary artery pressure was measured during mechanical ventilation after cardiac surgery in 53 patients; 37 of the patients were studied again after extubation. Three monitoring methods were compared: graphic strip recording, the "stop cursor" (monitor screen freezing) method, and digital-display recording. Difference scores were calculated between the methods and analyzed for frequency and direction. RESULTS: All comparisons showed differences of at least +/-3 mm Hg in measurements obtained with the three methods. During mechanical ventilation, the digital and graphic measurements of systolic pressure varied most often; 57% (30/53) of the comparisons had difference scores of at least +/-3 mm Hg. The cursor and graphic measurements of diastolic pressures varied least often; 6% (3/53) of the comparisons had difference scores of at least +/-3 mm Hg. As expected, the digital method most often gave higher results than the graphic method. During spontaneous breathing, measurements of systolic pressure varied more often (38% to 53%) than did measurements of diastolic pressure (12% to 37%). Unexpectedly, for systolic pressures, the difference between digital and graphic measurements was 3 mm Hg or more 30% (11/37) of the time, and the difference between cursor and graphic measurements was 3 mm Hg or more 53% (17/32) of the time. CONCLUSIONS: Because of physiological and technical influences, measurements of systolic and diastolic pressures in the pulmonary artery made with the digital and cursor methods were not as reliable as measurements made with the graphic method. The findings support continued use of the graphic method for accurate measurements of pulmonary artery pressure.