Peripheral polyneuropathies are common and their diagnosis may be challenging. We compared the results from sural-nerve and skin biopsies in 33 patients with a polyneuropathy and neuropathic pain examined in our hospital over a 6-year period. The biopsies were all from the same lower limb of each patient. Intraepidermal nerve fiber (IENF) densities in the skin were determined by fluorescence microscopy; unmyelinated fiber densities in sural-nerve biopsies (UFNB) were determined by electron microscopy. There was no correlation with age or gender in either biopsy type; there was a weak trend to correlation between UFNB density and IENF density, possibly because of the small sample size. The sensitivity of detection of quantitative abnormalities of unmyelinated fibers was better in the skin than in the nerves. Proximal and distal IENF densities were strongly correlated; and counts of UFNB were highly reproducible. Thus, quantification of unmyelinated fibers in sural-nerve and skin biopsies seem to be complementary. Sural-nerve biopsy may be required to confirm a specific diagnosis, to identify lesion mechanisms, and to devise therapeutic strategies, whereas skin biopsy seems to be more efficient in the follow-up of length-dependent polyneuropathies and in the diagnosis of neuropathic pain.