Background: Advanced cancer is associated with a significant symptom burden, and timely identification of palliative care (PC) needs, and provision of appropriate PC can improve treatment outcomes, reduce healthcare cost, and enhance patient and family satisfaction with care. Several tools have been used to identify PC needs in different clinical settings and patient groups.Objective: The primary objective was to determine the prevalence and associated characteristics of PC needs among cancer patients admitted to the surgical emergency center (SEC) of a large academic hospital in South Africa (SA). The association between PC needs and early death were explored as a secondary outcome.Design: This was a cross-sectional observational study that included all patients with known malignancy admitted through the SEC for acute surgical emergencies. The validated Supportive and Palliative Care Indicators Tool (SPICT™) was applied to patients' files on admission to the SEC. In addition, attending physicians were asked to estimate the 1 year survival probability of these patients by answering The Surprise Question (SQ).Setting: A tertiary level, public, academic hospital in Cape Town, SA.Results: One hundred and twelve admissions were included with a median age of 58 years. Fifty-two admissions (46.4%) were for metastatic patients and 60.7% were known with palliative treatment intent. The prevalence of SPICT- and SQ-defined PC needs was 46.4% and 54.7% respectively. Pain was the most prevalent presenting symptom and bowel obstruction the most prevalent presenting diagnosis. SPICT-positivity was a significant predictor of death before discharge and death within 6 months of first admission. Proportional agreement in predicting for PC needs of greater than 70% was shown between the two tools.Conclusion: Patients with PC needs comprise a significant proportion of SEC cancer admissions. This study shows the clear need for investment in staff and infrastructure to provide integrated palliative and end-of-life care as part of surgical services. The SPICTTM and SQ were shown to predict for early death in this cohort. Further validation of PC needs assessment tools is needed to guide the cost-effective implementation of PC services in low resource settings. [ABSTRACT FROM AUTHOR]