Various parameters of the repetitive nerve stimulation (RNS) test of the abductor digiti quinti muscle were analyzed statistically in 34 patients with Lambert-Eaton myasthenic syndrome (LEMS). The sensitivity and specificity of the increments after exercise and after 50-Hz stimulation for the diagnosis of LEMS were compared with reference values in 40 normal subjects and data from 538 tests in patients with myasthenia gravis (MG). When we used a 100% increment (the "gold standard") as the normal limit for the postexercise facilitation (PEF) or the high-rate stimulation (HRS) test, the diagnosis of LEMS was confirmed in 29 (85%) cases. When a 60% increment was used as the normal limit, the diagnosis of LEMS was made in 97% of cases. In MG, a 60% increment was observed in only 4 of 538 cases by HRS and in none by the exercise test. Thus, the use of a 60% increment showed a sensitivity of 97% for the diagnosis of LEMS and a specificity of 99% in excluding MG. A 60% increment in either the PEF or HRS test for the diagnosis of LEMS is a desirable alternative to the 100% increment previously considered to be the gold standard for this diagnosis. © 2005 Wiley Periodicals, Inc.