Objective Verbal episodic memory decline is the core feature of amnestic mild cognitive impairment (MCI) and early stage Alzheimers disease (AD), the delayed story recall (DSR) from Wechsler Memory Scale (WMS-Ⅲ) Logical Memory, Story A could significantly discriminate MCI from normal aging and predict its conversion to Alzheimers dementia.However, there are few studies about the DSR scale in Chinese, and also no report of the cut-off line of this subtest in Chinese-speaking population.Hence, aim of this study was to validate the DSR Chinese version and develop proper cut-off point in detecting MCI.Methods Total 1269 subjects were screened from 9 sites in China between January 2006 and April 2011.Of them, 443 subjects were considered as normal cognition (NC), and 578 were considered as MCI, 248 met the diagnosis of AD.According to the criteria for MCI proposed by Petersen et al., a formula [cutoff score =normal mean-(SD × 1.5)] was used to calculate the cutoff score of DSR scale for detecting MCI.The receiver operating characteristic analysis (ROC) was used to assess the ability of the DSR cutoff score.Results The mean DSR cutoff score with the optimal sensitivity (0.949) and specificity (0.882) was 11.8 for discriminating MCI from NC (Figure 1).Because the total DSR score was influenced by age(r =-0.096, P =0.001), so the cutoff score was calculated for age, in the 50 to 64 years old patients, when the cutoff score was 13.8 of DSR, the optimal sensitivity (0.941) and specificity (0.944) was obtained.An optimal sensitivity (0.960) and specificity (0.947) were obtained with the cutoff score were 10.4 of DSR in the 65 to 74 years old patients, and 9.8 of DSR in the 75 to 85 year old patients with the sensitivity (0.987) and specificity (0.877).The score of DSR was significantly correlated to the MMSE (r =0.498, P =0.000) and showed higher retest reliability (r =0.887, P =0.011) after a 3 month follow-up.Conclusion The DSR Chinese version can be used as a screening tool to detect MCI and AD with high sensitivity and specificity, and it should be used to select person with high risk of cognitive impairment.