The purpose of this study is to evaluate the reliability of motor nerve conduction velocity (MNCV) in the assessment of uremic neuropathy. Ulnar and peroneal MNCVs were measured by Hopf technique as well as conventional method in 19 patients with renal failure in conservative therapy (Group I), 15 patients with renal failure in dialysis therapy (Group II) and 32 healthy volunteers (Control). The results were as follows:

1) Most of the cases showed normal range values of motor conduction velocity by conventional method, but the patients with renal failure in conservative therapy showed negative relationship between the serum creatinine level and conduction velocity of ulnar nerve.

2) The mean maximal and minimal conduction velocity of ulnar nerve were 60.74±9.65 m/sec 43.15±7.94 m/sec in group I, 54.87±9.66 m/sec 38.17±7.45 m/sec in group II and 64.88±7.17 m/sec 49.65±4.89 m/sec in control group, respectively.

3) The mean maximal and minimal conduction velocity of peroneal nerve were 48.51±3.73 m/sec 37.40±4.58 m/sec in group I, 47.93±7.78 m/sec 34.43±6.96 m/sec in group II and 56.27±7.18 m/sec 42.11±5.88 m/sec in control group, respectively.

4) By comparing conventional method with Hopf collision method, the measurement of minimal conduction velocity using Hopf's technique (collision method) is more precise estimate in detecting of polyneuropathies which is not detected by conventional nerve conduction study.

5) The maximum likelihood estimate between renal failure groups and control group was noted when the minimal conduction velocity of 5.05 m/sec cutpoint was used for ulnar nerve (odds ratio 14.12, sensitivity 94%, specificity 47%, concordance rate 71%) and 43.0 m/sec for peroneal nerve (odds ratio 11.00, sensitivity 97%, specificity 25%, concordance rate 62%).