This study aimed to preliminarily explore the reliability and validity of the Chinese version of the Cognitive Assessment for Stroke Patients (CASP) in patients with nonaphasic stroke and provide a reliable basis for its clinical application in China. The original French version of the CASP was translated into Mandarin Chinese. The study enrolled 58 patients in the neurological center. Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), and CASP were used to evaluate cognitive function. Content validity, structural validity, and concurrent validity, internal consistency, interrater consistency, and retesting reliability were used to evaluate the results. The Spearman correlation coefficient of each item and the total CASP score was between 0.320 and 0.905 ( 0.5, indicating good construct validity. The total CASP score significantly correlated with the total MMSE ( = 0.796,  < .001) and total MoCA ( = 0.816,   0.85) and retesting reliability (ICC = 0.7-0.951) were good. The Chinese version of the CASP has good reliability and validity and can play a good auxiliary role in evaluating cognitive dysfunction in patients with stroke.