Adrenal masses are discovered in 5% of abdominal imaging scans and the diagnostic work-up has to address two major questions: does the adrenal mass overproduce hormones and, most important to the patient, does the adrenal tumour represent cancer? The reported incidence of adrenocortical carcinoma varies in reported series from 2 to 11%. However, the accuracy of currently available imaging tests to diagnose malignancy is poor. Urine steroid metabolomics is a combination of steroid profiling and machine learning-based data analysis, which was previously reported to diagnose adrenocortical carcinoma (ACC) with 90% sensitivity and specificity. We recently undertook a prospective international multi-center test validation study of urine steroid metabolomics, recruiting >2000 patients with newly diagnosed adrenal mass and employing high-throughput urine steroid metabolite analysis by tandem mass spectrometry prior to computational data analysis. Results demonstrate an excellent diagnostic performance of urine steroid metabolomics, superior to the performance of routine imaging, with an algorithm combining urine steroid metabolomics with two imaging parameters providing the best performance. We propose that urine steroid metabolomics should become part of the standard-of-care in the diagnostic work-up of patients with indeterminate adrenal tumors.