The high cost of postexposure prophylaxis for rabies is one reason that treatment is inadequate in developing countries. This problem has kindled interest in the use of equine rabies immune globulin, which is a less expensive, yet effective, substitute for human rabies immune globulin. Fatal anaphylaxis is a feared complication of the administration of heterologous serum; therefore, authoritative sources recommend prior skin testing. However, recommendations for methods of administering such a skin test and for its interpretation vary greatly. We embarked on a long-term study to develop guidelines for administration and interpretation of skin test results and to eventually determine a cost-benefit ratio. The skin test is not predictive of serum sickness. Anaphylaxis is rare with modern purified and pepsin-digested equine rabies immune globulins. We consider a skin test to be positive only if a wheal of > 10 mm in diameter, with or without flare, or a wheal of 5-10 mm in diameter with a flare of > 20 mm develops. Long-term studies will be required to answer the remaining questions regarding test criteria and cost benefits.