We have compared ultrasound characteristics of spread during infraclavicular brachial-plexus blocks by use of electrically evoked radial-nerve- or median-nerve-type distal motor responses to guide the injection of 30 mL of 1.5% mepivacaine. Consecutive patients who required surgery distal to the upper arm were prospectively included in this study. With radial- or median-evoked distal motor response at a stimulating current intensity of less than 0.5 mA, patients were distributed into 2 equal groups. An independent investigator blinded to the evoked response described ultrasound characteristics of the spread of local anesthetic and assessed block quality 30 minutes after placement. A quality diffusion score proportional to the extent and intensity of spread around the axillary artery was used, and dynamic movements during injection were noted. Thirty-two patients were included. With radial-nerve-type motor response, the success rate of infraclavicular plexus block was 100%, but 3 supplemental axillary blocks were requested with median-nerve-type motor response. Quality diffusion scores were significantly higher with radial-nerve-type as compared with median-nerve-type motor response (P = .03). Injection after radial-nerve-type motor response resulted in a typical and reproducible ultrasound feature of posterior local-anesthetic spread associated with medial and upper movement of the axillary artery. With median-nerve-type motor response, failed blocks were associated with a specific posterior displacement of the axillary artery that resulted from superficial spread. We have demonstrated that as compared with median-nerve-type motor response, injection performed after a radial-nerve-type motor response promoted reproducible and remarkable ultrasound spread characteristics associated with complete sensory block of the 3 cords at 30 minutes.