Background: Approximately 20% of women receiving endocrine therapy (ET) for early-stage breast cancer (BrCa) discontinue their regimen due to side effects, including vasomotor symptoms and arthralgias, subjecting them to higher rates of disease recurrence. Exercise is one approach to mitigate ET-related toxicity and potentially improve medication adherence by mitigating ET-related symptoms. However, only 11% of women with BrCa adhere to recommended exercise guidelines. Purpose: To examine the effect of a virtual supervised exercise program for women receiving ET for early-stage BrCa on physical function (PF) (primary outcome) and adherence to ET (secondary outcome). Methods: Participants were recruited through referral by oncology care providers from multiple BC Cancer sites and self-referral. Eligible participants were: females with stage I-III BrCa being treated with curative intent, 19-75 years, taking ET for at least 3 months up to 3 years. Using a partial cross-over design (NCT04824339), participants were randomized to the immediate exercise group (INT) or the delayed exercise group (CON), which crossed over at 8 weeks. The 8-week intervention was comprised of a 50-minute resistance exercise circuit, delivered virtually by an exercise physiologist twice per week via Zoom, and self-directed aerobic exercise (progressing to 90 minutes or more per week of at least moderate intensity). PF was assessed using the 30-second chair stand test. ET adherence was measured by the Voils self-reported medication non-adherence questionnaire. Values reported as means and standard deviations. The difference at 8 weeks by intervention arm was examined using two sample t-test (continuous) or kappa measure of agreement (categorical). Results: A total of 132 women were randomized. On average, women were 54 years, White (European) (47%), married (64%), had at least a bachelor’s degree or higher (51%), and had a body mass index of 27.7 kg/m². The most common type of ET women received was Letrozole (43%), followed by Tamoxifen (36%). At 8 weeks, there was a significant improvement in PF for INT versus CON (mean change +2.5±2.9 vs. +0.7±2.2 stands, p< 0.001). There was no difference between groups in change in medication adherence status (k = 0.35). Conclusions: An 8-week virtual supervised exercise program for women receiving ET for early-stage BrCa significantly improved PF but did not change medication adherence at 8 weeks. The potential impact of the intervention on long-term medication adherence will be explored up to one year and with covariates. Based on improvement in PF, future trials to establish effectiveness and inform implementation are warranted. Clinical trial information: NCT04824339.