This study examines current criteria for CIDP by comparing the electrodiagnostic data from 17 patients who have biopsy proven CIDP with 29 patients with diabetic neuropathy. The groups were comparable in age. Data were examined using four sets of published criteria for primary demyelination. Two sets (A & B) defined demyelination based on distal motor latencies, F-wave latencies, motor conduction velocities, and temporal dispersion or conduction block. The other two sets defined demyelination based on slowing of motor conduction velocity < 70% of the lower limit of normal (LLN) in either 2 nerves (set C), or in a single nerve (set D). Using set A, 3/17 patients with CIDP met the criteria for a demyelinating neuropathy, while this was true in 2/29 of the patients with diabetic neuropathy. For sets B and C, data consistent with a demyelinating neuropathy were identified in 2/17 and 1/29 patients, respectively. Using set D criteria for demyelination was met in 6/17 of the patients with CIDP and in 5/29 of the diabetic patients. None of the criteria sets showed a significant difference between CIDP and diabetic neuropathy (p = 0.3 to 0.5). Though electrodiagnosis is an important element in the diagnosis of CIDP, current electrodiagnostic criteria alone are insufficient for defining many cases of CIDP.