AimThis study explored the possibility of developing an alternative, simple and rapid test for the screening of distal symmetrical polyneuropathy (DSPN), for use in local primary care facilities. Methods4023 patients with type 2 diabetes mellitus (T2DM), aged 18–80 years, from 21 hospitals in China were screened for inclusion in a prospective diagnostic assessment of DSPN, between April 2011 and March 2012. Patients underwent nerve conduction study (NCS; as the gold standard) and clinical assessments of ankle reflexes (AR) and thermal (T), vibration (V), pressure (P), and pinprick (PP) sensations. ResultsResults from various combinations of these clinical assessments were compared to determine the optimal screening regimens. The Toronto Consensus Panel definition was used to confirm clinical DSPN (symptoms or signs plus abnormal NCS). 3883 of the 4023 patients (M:F, 2162:1721; mean age, 56.79±13.16 years) were included. No significant differences were found between the two most efficient clinical assessment combinations, AR+PP+T+V and AR+T+V (P=0.9612). Compared to the diagnosis made using AR+T+V, the same diagnosis was reached using AR+PP+P+V+T in 100% of patients with probable clinical DSPN, 84.05% with possible clinical DSPN, and 96.84% with no DSPN. The AR+T+V assessment was also ∼40% faster than the AR+PP+P+V+T assessment. ConclusionsCombined AR+T+V assessment is a valuable tool for screening diabetic patients for DSPN, potentially decreasing costs to local primary care facilities and making care more available to underserved populations.