BACKGROUND: Detecting urinary Bence Jones protein (BJP) is effective for early diagnosis of malignant diseases, such as multiple myeloma. Currently, BJP is detected by heat coagulation test (Putnam's method), immunoelec-trophoretic (IEP) analysis, and immunofixation electrophoresis (IFE). However, because of their low sensitivity, high cost, and labor intensiveness, alternative biomarkers are being sought. The study aimed to evaluate the use-fulness of the urinary albumin/total protein (U-ALB/U-TP) ratio for early detection of urinary BJP. METHODS: U-ALB and U-TP were measured in 85 BJP urine samples confirmed by IEP or IFE and 507 samples from the general patient group, for a total of 592 samples. A total of 592 general patient samples were used to calculate the U-ALB/U-TP ratio for receiver operator characteristic (ROC) curve analysis. RESULTS: The correlation between U-TP and U-ALB was strong in the general group but weak in the BJP group. For a BJP detection cutoff value obtained by ROC curve analysis of 0.362 (area under the curve: 0.941), the sensitivity was 92% and the specificity was 88% for a U-TP ≥ 30 mg/dL. A decreased ratio of U-ALB/U-TP indicates turbid urine (high white blood cell count), bacteriuria, crystalluria, or high urinary free L chain as well as urinary BJP. CONCLUSIONS: We demonstrated that BJP could be detected by measuring the U-ALB/U-TP ratio, which may enable early detection of BJP in patients with malignant diseases at the presymptomatic stage.