The electrodiagnostic yield of the medial plantar nerve action potential (NAP) amplitude versus the sural/radial amplitude ratio (SRAR) was determined in 110 consecutive patients with clinically diagnosed distal sensory polyneuropathy (SN) and normal sural responses. Forty-five consecutive patients with clinically diagnosed lumbosacral radiculopathy served as disease controls. Of the 110 SN patients, 32 were classified clinically as SN with large-fiber involvement (SN-LFI), whereas 78 had clinically pure small-fiber SN. Plantar NAP amplitudes were abnormal in 18 of 32 patients (56%) with SN-LFI, and 15 of 78 (19%) with small-fiber SN. A SRAR <0.21 (fifth percentile of normal) was found in 7 of 32 patients (22%) with SN-LFI and 8 of 78 (10%) with small-fiber SN. In the control group, the medial plantar NAP was normal in all 45 subjects (100%), whereas the SRAR was >0.21 in 43 subjects (96%). Thus, for a 50% pretest probability of SN-LFI, the positive predictive value of an abnormal medial plantar was 100% versus 85% for a SRAR <0.21. The medial plantar NAP amplitude is a more useful measure of SN, than is the SRAR, in patients under age 70, with suspected SN-LFI. The yield of the SRAR and plantar NAP amplitude is poor when clinical signs of large-fiber sensory dysfunction are lacking.