Background: Patients with cancer can experience disease- and treatment-related symptoms that are underreported and underestimated by physicians. This observational, non-interventional study evaluated the use of ePROs and their impact on duration of treatment (DoT) in patients with solid tumors receiving IO therapy in community practice. Methods: Patients initiating index IO therapy immediately prior to (Jan-2017 to Dec-2018) and after (Sep-2019 to Dec-2020) implementation of Noona, the ePRO platform at Tennessee Oncology clinics, were included in a retrospective historical control (HC) and ePRO cohort, respectively, and followed for up to 6 months. The ePRO cohort was further divided into ePRO users (platform enrollment ≤45 days from index) and non-users. ePRO questionnaires, based on Common Terminology Criteria for Adverse Events (CTCAE), were sent within a week after each IO infusion and could be completed using an internet browser or smartphone app. Patient characteristics and DoT were described and compared between the HC and ePRO cohorts and between the HC cohort and ePRO users subgroup. Use of ePROs was evaluated within the ePRO cohort. Differences in baseline characteristics between cohorts were adjusted using Cox proportional hazards models. Results: Data were collected for 538 HC and 1014 ePRO patients (319 ePRO users and 695 non-users). Patient characteristics were generally similar between cohorts, but more HC patients were diagnosed with Stage IV disease (54% vs 47%; p < 0.01) and initiated IO as monotherapy (82% vs 52%), while more ePRO patients initiated IO as combination therapy (48% vs 18%). ePRO users were more likely than non-users to be female, white, married, living with a spouse, and have higher education (college or graduate degree) (all p < 0.05). Use of ePROs was durable over follow-up, with a consistent number of questionnaires sent over Months 1-3 and Months 4-6 (median: 6 questionnaires in each period) and a slight decrease in the number answered (median: 4 vs 3 questionnaires). ePRO patients had a longer DoT than HC patients (median time to end of first IO regimen: not estimable vs 5.1 months). Significantly more ePRO than HC patients remained on their first IO regimen at 6 months (54% vs 46%; p < 0.05). Multivariate Cox regression showed the risk of ending first IO therapy was lower for ePRO versus HC patients (p < 0.05). Conclusions: The increased DoT observed in the ePRO versus HC cohort in this study suggests that use of ePROs may facilitate improved care coordination and enable patients to remain on IO therapy longer. However, ePRO uptake was only 31% in the ePRO cohort, with several social determinants appearing to influence use. Overcoming barriers in ePRO uptake is an area for future study.