Residual neuromuscular blockade can be evaluated using acceleromyography, tactile assessment of train-of-four (TOF), double-burst stimulation (DBS), 50-Hz tetanus, or 100-Hz tetanus. Nerve stimulation can be at the hand or the wrist. We compared all these tests at both sites of stimulation. Rocuronium was given to 32 patients under sevoflurane anesthesia. The mechanomyographic adductor pollicis TOF ratio was measured at one extremity. In the other, stimulation was at the hand or the wrist, by random allocation, and the acceleromyographic TOF ratio was measured. During recovery, a blinded observer estimated tactile fade. The TOF fade became undetectable when mechanomyographic TOF ratio was (mean ± sd) 0.31 ± 0.15. For DBS, this threshold was 0.76 ± 0.11. For 50-Hz tetanus, it was 0.31 ± 0.15. For 100-Hz tetanus, it was 0.88 ± 0.18, with a range of 0.14–1.00. These tactile responses were the same for hand and wrist stimulation. When acceleromyographic TOF ratio reached 1.0, the mechanomyographic TOF ratio was 0.89 ± 0.06. With stimulation in the hand, acceleromyographic TOF ratio >1.0 was less frequent than at the wrist. To exclude residual paralysis, TOF, DBS, and 50-Hz tetanus are inadequate, 100-Hz tetanus is unreliable, and acceleromyography performs best.