Purpose Although the Supportive Care Needs Survey is one of the most comprehensive and robust cancer-specific needs assessment instruments, no version exists specifically for cancer patients in mainland China. This study tested the psychometric properties of the Mandarin version of the 34-item Short-Form Supportive Care Needs Survey ( SCNS-SF34-C (Mandarin)) in mainland Chinese cancer patients. Methods From December 2015 to May 2016, patients were recruited from two cancer centers in Guangzhou, China, to complete the SCNS-SF34-C (Mandarin). Exploratory factor analysis (EFA) was used to test the factor structure. The internal consistency, convergent validity, and discriminant validity of the resulting factor structure were evaluated by traditional psychometric analysis. Results A total of 861 patients completed the SCNS-SF34-C (Mandarin). Item 14 was removed for its low factor loadings on every factor in the initial EFA. Using the remaining 33 items, the reiterated EFA produced a five-dimension structure that was consistent with the dimensions of the original version of the SCNS-SF34 (health system and information, psychological, patient care and support, physical and daily living, and sexuality), accounting for 69.757% of the total variance. Cronbach's alpha coefficients ranged from 0.854 to 0.942 for the five domains and 0.947 for the whole scale. Convergent validity was verified by significant correlations with all corresponding instruments. It discriminated between groups based on age, sex, marital status, and stage of disease. Conclusions Preliminary evidence suggests that the SCNS-SF34-C (Mandarin) is a reliable and valid instrument for assessing the supportive care needs of cancer patients in mainland China.