OBJECTIVE To determine the feasibility of examining intraepidermal nerve fiber density (IENFD) in postmortem skin. METHODS From 12 subjects, 3-mm skin punch biopsies were collected 1-4 days postmortem from the proximal leg and distal leg, with a mean (range) interval from the death of 37 (15-91) hours. Causes of death varied broadly, including hepatocellular carcinoma, chronic lymphocytic leukemia, generalized atherosclerosis, progressive supranuclear palsy, Parkinson disease, emphysema, and obesity. The mean (range) number of sections evaluated from each biopsy was 5.08 (2-6) from the proximal leg and 5.92 (5-6) from the distal leg. Sections were stained with PGP 9.5 for blinded counting using bright field microscopy. Qualitative and quantitative assessment of feasibility included a comparison of fiber staining with that in healthy subjects and mean IENFD in postmortem samples. Interobserver reliability was assessed among 3 blinded raters by calculating intraclass correlation coefficients and percentage variability of IENFD in at least 4 sections from biopsies in 5 healthy subjects. RESULTS Intraobserver and interobserver correlation coefficients of blinded IENFD counts undertaken by 4 authors were consistently >0.80, and the coefficient of variation was ≤10%. The quality of staining in postmortem samples was comparable with that in healthy subjects and was not substantially affected by time from death to specimen collection of up to nearly 4 days. Mean (range) IENFD from postmortem samples in the proximal and distal leg was 2.73 (0-7.65) and 1.93 (0-4.91) fibers/mm of skin, respectively. Two of 3 patients who had received chemotherapy during life showed a nearly complete absence of intraepidermal nerve fibers. CONCLUSIONS IENFD measurement in postmortem skin is feasible and may be used to study the epidemiology of SFN.