Background: Young men who have sex with men and transgender women (YMSM/TGW) have disproportionately high HIV incidence and lower preexposure prophylaxis (PrEP) adherence. Point-of-care (POC) urine tenofovir (TFV) rapid assay (UTRA) testing permits real-time monitoring for nonadherence within clinical settings. We performed UTRA testing among PrEP users to examine the relationship between low PrEP adherence and future PrEP discontinuation, and the accuracy of POC testing compared to gold-standard liquid chromatography tandem mass spectrometry (LC/MS/MS). Methods: YMSM/TGW participants ( n = 100) were recruited during a daily PrEP visit. Logistic regression models analyzed the relationship between the primary predictor of urine POC assay results (cutoff 1,500 ng/ml) and the primary outcome of PrEP discontinuation, defined as no PrEP follow-up or prescription within 120 days. Results: Overall, 19% of participants had low urine TFV and 21% discontinued PrEP, while 11% of participants self-reported low PrEP adherence (<4 pills per week), which was only 43% sensitive/84% specific in predicting low TFV levels and was not associated with PrEP discontinuation. Low urine TFV level predicted PrEP discontinuation adjusted odds ratio (AOR) 6.1; 95% confidence interval (CI): 1.4–11; P = 0.005 and was 71% sensitive/90% specific for discontinuation after 120 days. Compared to LC/MS/MS, UTRA testing had a 98% positive and 100% negative predictive value. Conclusions: In a sample of YMSM/TGW on daily PrEP, POC UTRA testing predicted PrEP discontinuation more accurately than self-reported adherence, with high predictive values compared to LC/MS/MS. UTRA testing may be a clinical tool for directing preventive interventions towards those likelier to discontinue PrEP despite ongoing HIV vulnerability.