Two techniques for blocking the lumbar plexus were prospectively evaluated in 50 children undergoing surgery in the hip region and randomly allocated to one of two equal groups. A variant of the "psoas compartment block" and the classic technique were used in groups 1 (n = 25) and 2 (n = 25), respectively. All procedures were carried out under light general anesthesia with the patients in the lateral position using insulated needles and electrical stimulation. Both procedures were effective, allowing completion of surgery without additional treatment in almost all patients. However, the distribution of analgesia differed: 23 (ipsilateral) lumbar and sacral plexus blocks and 2 (ipsilateral) lumbar blocks alone were produced in group 2, compared to 22 areas of anesthesia comparable to those that might be associated with a lumbar epidural block and two ipsilateral lumbar plexus blocks in group 1. The two techniques are not, therefore, mere variants of the same basic approach to the lumbar plexus. The procedure described by Winnie et al. (Anesthesiol Rev 1974;1:11-6) was more suitable for providing unilateral blockade than the "psoas compartment block."