Recording the tibial H reflex from the calf is routine in clinical electromyography. Two widely used methods are Hugon's and Braddom and Johnson's. Both methods produce an H wave with an initial positive deflection. Another method, the "half and half" method, generates an H wave with an initial negative deflection. To compare these methods, H-reflex responses were recorded in the right calf of 20 normal adults with each method using identical amplifier and stimulator settings. Stimulus current was increased by increments of 0.5-1.0 mA until the H reflex and motor (M) wave responses reached maximum amplitudes (Hmax, Mmax). H wave to M wave (H/M) amplitude ratios were calculated. The half-and-half method produced greatest Hmax, Mmax, and H/M ratio. Wave form configuration for the half-and-half method was biphasic with an initial negative deflection; it was triphasic with an initial positive deflection with the other methods. The shapes of M and H waves were similar in two methods but dissimilar in one. Because of the greater amplitude and the initial negative deflection, the H wave may be easier to detect with the half-and-half method and may present a latency that is more easily determined. For these reasons, the half-and-half method is recommended for measuring amplitude and latency in clinical settings.