We examined a physician-performed, goal-directed ultrasound protocol for the emergency department management of nontraumatic, symptomatic, undifferentiated hypotension. Randomized, controlled trial of immediate vs. delayed ultrasound. Urban, tertiary emergency department, census >100,000. Nontrauma emergency department patients, aged >17 yrs, and initial emergency department vital signs consistent with shock (systolic blood pressure 1.0), and agreement of two independent observers for at least one sign and symptom of inadequate tissue perfusion. Group 1 (immediate ultrasound) received standard care plus goal-directed ultrasound at time 0. Group 2 (delayed ultrasound) received standard care for 15 mins and goal-directed ultrasound with standard care between 15 and 30 mins after time 0. Outcomes included the number of viable physician diagnoses at 15 mins and the rank of their likelihood of occurrence at both 15 and 30 mins. One hundred eighty-four patients were included. Group 1 (n = 88) had a smaller median number of viable diagnoses at 15 mins (median = 4) than did group 2 (n = 96, median = 9, Mann-Whitney U test, p <.0001). Physicians indicated the correct final diagnosis as most likely among their viable diagnosis list at 15 mins in 80% (95% confidence interval, 70-87%) of group 1 subjects vs. 50% (95% confidence interval, 40-60%) in group 2, difference of 30% (95% confidence interval, 16-42%). Incorporation of a goal-directed ultrasound protocol in the evaluation of nontraumatic, symptomatic, undifferentiated hypotension in adult patients results in fewer viable diagnostic etiologies and a more accurate physician impression of final diagnosis.