BC diagnosis relies on the insertion of an optical endoscope into the bladder cavity through urethra to image the suspected lesions. [7,8] This process is highly invasive and would cause urethra and bladder injury, resulting in hematuria and even urinary bacterial infection within a few days after examinations. [9] Critically, cystoscopy diagnosis is plagued with the inherent bladder tumor heterogeneity, and therefore, limits the accuracy of early BC diagnosis. [10] Recently, noninvasive liquid biopsy emerges as an alternative to address the bottleneck of spatiotemporal tumor heterogeneity and to obtain disease-relevant molecular information for clinical cancer diagnosis and cancer status monitoring. [11][12][13][14][15][16][17][18][19][20] Bladder is a urine storage organ that has been recognized as the metabolic microenvironment for bladder tumor cells, and its carcinogenesis and progression could make a pivotal impact on urine. [21] In this regard, the development of a urine biopsy provides a powerful strategy toward noninvasive early diagnosis and prognosis of BC. Clinically, urinalysis has been routinely utilized to detect abnormal metabolic biomarkers in urine for the assessment of health status and preliminary screening of diseases. [22][23][24][25] However, the physiologically relevant biomarkers detected by routine urinalysis are generally limited to high concentration targets over the micromolar level. On the other hand, the concentration Urinalysis is attractive in non-invasive early diagnosis of bladder cancer compared with clinical gold standard cystoscopy. However, the trace bladder tumor biomarkers in urine and the particularly complex urine environment pose significant challenges for urinalysis. Here, a clinically adoptable urinalysis device that integrates molecular-specificity indium gallium zinc oxide field-effect transistor (IGZO FET) biosensor arrays, a device control panel, and an internet terminal for directly analyzing five bladder-tumor-associated proteins in clinical urine samples, is reported for bladder cancer diagnosis and classification. The IGZO FET biosensors with engineered sensing interfaces provide high sensitivity and selectivity for identification of trace proteins in the complex urine environment. Integrating with a machine-learning algorithm, this device can identify bladder cancer with an accuracy of 95.0% in a cohort of 197 patients and 75 non-bladder cancer individuals, distinguishing cancer stages with an overall accuracy of 90.0% and assessing bladder cancer recurrence after surgical treatment. The non-invasive urinalysis device defines a robust technology for remote healthcare and personalized medicine.