Purpose <p>Clinician-assessed performance status is a strong predictor of morbidity in cancer patients, but it lacks patients’ perspective. This study aims to examine the agreement between clinician (cPS) and patient (pPS) in reporting performance status and identify which assessment better predicts 30-day morbidity.</p> Methods <p>This was a prospective observational study among patients undergoing the first cycle of cytotoxic chemotherapy for solid tumour malignancies at Malabar Cancer Centre, Thalassery, Kerala, India, from February 1 to August 31, 2024. Patient and clinician versions of the ECOG PS questionnaire were used to get pPS and cPS assessments. Average PS was calculated as the mean of absolute values of cPS and pPS. Patients were treated as per standard protocol. Outcome variables, including 30-day morbidity, hospitalization, and Grades 3–4 toxicities, were monitored over a 30-day follow-up period.</p> Results <p>Total 215 patients were included. Clinician and patient had only weak agreement in reporting performance status, with 14.65% agreements and weighted kappa = 0.042 (95% CI − 0.028 to 0.041). The mean pPS was slightly higher than mean cPS (1.62 vs 1.31). Only cPS had statistically significant association with 30-day morbidity and occurrence of toxicities. Hence, cPS had better odds of predicting 30-day morbidity and occurrences of toxicities compared to pPS. Average PS also had statistically significant association with 30-day morbidity, occurrence of hospitalisation, and Grade 4 toxicities.</p> Conclusion <p>Clinician-reported performance status remains the strongest predictor of chemotherapy-related morbidity and toxicities. The novel “Average PS” is simple and easily derivable score, incorporating the patient’s perspective, which needs validation in larger studies.</p> Trial registration <p>This study was registered in Clinical Trial Registry-India (CTRI/2024/03/064160).</p>