Objective: The present study aimed to validate a severity cut-off of negative symptoms for persistent negative symptoms (PNS) identification using the Clinical Assessment Interview for Negative Symptoms (CAINS). Method: A total of 206 patients with schizophrenia were recruited and divided into the PNS group (n = 57) and the Non-PNS group (n = 149) using PNS criteria based on the SANS and the SAPS. To determine the appropriate cut-offs on the CAINS in identifying PNS, Receiver Operating Characteristic (ROC) curve analysis was conducted in the PNS and Non-PNS groups. Results: Our results showed that the cutoffs for identifying PNS on the CAINS total score, the Motivation and Pleasure (MAP) subscale score and the Expression (EXP) subscale score were 25, 17, and 5 respectively. Area Under the Curve (AUC) analysis indicated excellent discrimination of the PNS group from the Non-PNS group using the cut-off for the CAINS total score. However, discrimination was somewhat better for the MAP subscale score than the EXP subscale score. The Positive Predictive Value (PPV) and Negative Predictive Value (NPV) of the MAP subscale were 81.54% and 97.16%. Conclusion: We found that the cut-off scores derived from the CAINS to identify PNS are comparable to existing scales. The CAINS offers an alternative means in identifying PNS patients in clinical trials that overcomes methodological and conceptual limitations of older scales. (C) 2018 Elsevier B.V. All rights reserved.