Background: Adjuvant endocrine therapy (AET) use among women with early-stage, hormone receptor-positive breast cancer reduces the risk of cancer recurrence, but adverse symptoms contribute to lower adherence. We evaluated the effectiveness of an app that integrates patient-reported outcomes with electronic health records and tailored messages. Methods: Women with early-stage breast cancer initiating AET (November 2018-June 2021) were randomized into three arms: (1) an “App” group that received instructions and access to the study app and weekly text reminders to use it; (2) an “App+Feedback (AF)” group received additional weekly tailored messages about managing symptoms, adherence, and communication; or (3) a “Usual Care (UC)” group without access to the app. The intervention lasted 6-months and participants completed surveys at enrollment, 6-, and 12-months. Increasing/severe symptoms and missed doses triggered alerts that prompted follow-ups from the oncology team. Outcomes included AET adherence, captured using an electronically monitored pillbox and self-reported adherence, symptom burden (FACT-ES), mental and physical health quality of life (SF-12), self-efficacy for managing symptoms (PROMIS 1.0), and 6-month count of emergency department visits/urgent care/hospitalizations [higher cost encounters] and office visits. Results: Overall, 300 women were randomized (102 UC, 96 App, and 102 AF). Median age was 60 years (range: 31–83), 34% identified as Black, 21% had incomes below 200% of the federal poverty level, and 20% had a high school degree or less education. Retention at 12-months was 88% (N=264). Mean app logins were 14.4 among intervention participants, which resulted in 4.2 mean alerts, with 58.0% having at least one alert during the 6-month intervention. AET adherence over 12-months measured using the pillbox was similar across groups: 76% for UC, 73% for App, and 71% for AF (p=0.57). At 12-months, AF participants had fewer higher cost care encounters over the previous 6 months (0.30, 95%CI: 0.15 to 0.45, p=0.01) compared to UC (0.71, 95%CI: 0.44 to 0.97) but not for App only vs. UC (0.46, 95% CI: 0.25 to 0.66, p=0.15). Mean mental health scores were 2.0 points better (95% CI: 0.21 to 3.84) at 6-months for AF (43.0) compared to UC (41.0). There were no statistically significant differences by group in self-reported adherence, symptom burden, physical health, self-efficacy for managing symptoms, and office visits at 6- or 12-months. Conclusions: While the intervention did not improve AET adherence, the app combined with tailored messages resulted in better self-reported mental health during the intervention and fewer higher cost care encounters at 12-months without increasing office visits. Symptom monitoring apps with tailored messages may be scalable and effective strategies for improving outcomes for patients with breast cancer. Clinical trial information: NCT03592771.