The authors studied the optimal epinephrine content of an epidural test dose, and determined criteria to identify intravascular injections in subjects with or without β-adrenergic blockade. Nine healthy nonpregnant subjects 25–36 years of age were given intravenous infusions of saline or esmolol in random order. During each infusion, they received a series of five injections (3 ml each) of either saline, 1% Iidocaine or 1% lidocaine containing 5, 10, or 15 μg of epinephrine. Thirty minutes after completing these two infusions, propranolol was administered as a bolus injection, and the series of five injections repeated. All injections were double blind and randomized. During saline infusion, all injections containing epinephrine significantly increased heart rate (HR) by an average of 31–38 beats/min when compared with that following plain lidocaine (P< 0.05), and increased systolic blood pressure by an average of 17–26 mmHg (P< 0.05 for the 15-μg dose only). During esmolol infusion, epinephrine injections increased HR by an average of 23–31 beats/min (P< 0.05), and increased systolic blood pressure by an average of 18–30 mmHg (P< 0.05 for 10 and 15 μg). After propranolol injection, epinephrine injections caused a decrease in HR by an average of 21–28 beats/min (P< 0.05), whereas systolic blood pressure increased by an average of 22–35 mmHg (P< 0.05 for 10 and 15 μg only). Without β-adrenergic blockade, an increase in HR ≥ 20 beats/min was 100% sensitive and specific for intravascular injection of 10 or 15 μg of epinephrine. After selective and nonselective β-adrenergic blockade, HR changes were not reliable, but an increase in systolic blood pressure (SBP) of ≥ 15 mmHg was diagnostic of injection of 10 or 15 μg of epinephrine. Hemodynamic changes occurred within 2 min after intravascular (iv) injections, and lasted at least 35 s. Injection of 5 μg of epinephrine did not produce reliable hemodynamic changes in any group. The authors conclude that in young, nonpregnant, individuals: 1) a test dose containing at least 10 μg epinephrine is a reliable marker of intravascular injection; 2) intravascular injection can be reliably detected by an absolute variation in HR of ≥ 20 beats/min in non–β-blocked subjects, and by an increase in SBP of ≥ 15 mmHg in the presence of β-adrenergic blockade.