The objective of this study was to determine the following: (1) if standard clinical evaluation is sufficient to provide an accurate estimate of hemodynamic status of unstable ICU patients; (2) the impact of pulmonary artery catheterization (PAC) on diagnosis and treatment plan; and (3) whether therapy provided after PAC was appropriate as judged by an expert panel of senior ICU physicians. A descriptive analysis of utilization of pulmonary artery catheters in a medical/surgical ICU population was performed in a university-affiliated hospital (24-bed medical/surgical ICU). The subjects included 154 medical/surgical patients judged by ICU residents and attendings to require PAC. All 154 patients underwent PAC with four patients having more than one catheterization. Prior to insertion of the catheter, a questionnaire was completed by medical/surgical residents and attendings indicating reasons for PAC insertion and estimate of hemodynamics. Following PAC, residents/attendings indicated their evaluation of hemodynamics and planned therapy. An expert panel rated performance of the house staff regarding treatment plan on a scale of 1 to 5 (5 indicating optimal therapy). The overall proportion correct classification for pulmonary artery wedge pressure (PAWP), CO, and systemic vascular resistance (SVR) were 47 percent, 51 percent, and 36 percent, respectively. In 45 percent of PAC, information obtained resulted in a major change in therapy. Major change in therapy occurred more often when prediction of PAWP by residents proved inaccurate. The expert panel judged appropriate scores of 3, 4 and 5 in 84 percent of the cases. Prediction of hemodynamics in ICU patients by clinical evaluation alone is inaccurate and unreliable. There is a positive correlation between inaccurate prediction of hemodynamics and major therapeutic changes after PAC. Most resident/attending performance was judged appropriate. Results of this study suggest that PAC was instrumental to the management scheme in many patients unresponsive to initial therapy. However, a subset of ICU patients were judged to have been managed favorably, yet had treatment based on inaccurate hemodynamic assessment.