This study evaluated the effects of cisplatin on 18 nerve conduction parameters of median, ulnar, peroneal, and sural nerves in relation to age, magnesium nadir, average magnesium, and magnesium replacement in gynecologic oncology patients. The 37 patients in this study received cisplatin (70 mg/m2) at 4-week intervals either as a single agent (17 patients) or in combination with doxorubicin and cyclophosphamide (20 patients) following inpatient hydration. The patients were placed randomly on either magnesium supplementation (intravenous plus oral) or placebo. For all patients, nerve conduction studies were performed before and after cisplatin therapy in the same EMG laboratory. Statistical analysis revealed that postcisplatin nerve conduction parameters were significantly predictable based upon pretherapy nerve conduction parameters, magnesium supplementation, total cisplatin, age, magnesium nadir, and average magnesium. Total cisplatin, age, and serum magnesium level were significant predictors in 8, 2, and 2 of the 18 nerve conduction parameters, respectively. Following cisplatin therapy, there was a significant decrease in sensory nerve action potential amplitude of median, ulnar, and sural nerves, whereas sensory latency of median and ulnar nerves was significantly increased. Cisplatin therapy had no effect on motor nerve conduction parameters of median, ulnar, and peroneal nerves. Factors responsible for the decrease in sural sensory action potential amplitude were not identified. Sensory nerve action potential amplitude and sensory latency of ulnar nerve are the two best objective parameters that can be utilized to monitor patients for adverse nerve conduction side effects of cisplatin.