We developed a TR-FIA kit for quantitative detection of CA50. This study aims to evaluate the analytical and clinical performances of this kit. Precision, accuracy, specificity, sensitivity, stability, and endogenous interference of this kit are evaluated. Reference range is established. Coincidence rate and correlation between TR-FIA and RIA are evaluated. ROC is adopted to evaluate the diagnostic performance. This kit shows excellent precision with a coefficients of variation (CVs) ranged from 2.2-9.3%, accuracy (average recovery, 98.5%), sensitivity (minimum detectable concentration is 0.2 U/mL), specificity (all cross-reactivity is less than 0.1% except CA199, which is 0.175%), and storage stability (recoveries, 90.8-100.4%). Bilirubin, hemoglobin, and triglyceride dose not interfere with CA50 detection (recovery, 97.13-109.1%). The range from 0-25 U/mL is chosen as the reference range. There are good correlation (r = 0.804) and coincidence (p = 0.608, kappa = 0.924) between TR-FIA and RIA. Diagnostic performance of this kits, which based on RIA results, is perfect (AUC = 0.996), and the diagnostic accuracy for malignancy diagnosis is in moderate degree (AUC, 0.802-0.861). The TR-FIA (CA50) kit performs well in analytical and clinical performances, and can be employed in the clinical diagnosis of malignancy.