The authors completed a two-phase study to determine criteria that might predict hemodynamic stability during recovery from subarachnoid block (SAB). Patients' supine and sitting (2 min) blood pressures were determined at 30-min intervals in the recovery room (RR). In the first group of 26 patients, retrospective analysis revealed that the orthostatic decrease in mean arterial pressure (MAP) never exceeded 15% following two successive orthostatic decreases of 10% or less. This finding was validated prospectively in a second group of 26 patients. Following two successive orthostatic MAP decreases of 10% or less, none of 65 orthostatic challenges resulted in an MAP decrease of more than 15%; conversely, in the absence of two successive MAP decreases of less than 10%, 5 of 51 orthostatic challenges resulted in an MAP decrease of greater than 15% (P less than 0.02). Had patients been discharged from the RR based on two successive MAP decreases of less than 10%, 35 of 52 patients could have been discharged from the RR 76 +/- 6 min (mean +/- SE) sooner than they would have under usual empirical discharge criteria of supine hemodynamic stability, regression of sensory level to T10, and return of toe movement. Following SAB, hemodynamic stability may return before sensory and motor function; for many patients, orthostatic testing following SAB may safely decrease the amount of time spent in the RR.