Background: We previously demonstrated the safety of stereotactic body radiotherapy followed by pembrolizumab (SBRT+P) in patients with advanced solid tumors. This phase 1 clinical trial was expanded to study the safety of partial tumor irradiation (partial-Rx). We assessed irradiated local failure (LF) and clinical outcomes with correlations to biomarkers including CD8-T-cell radiomics score (RS) and circulating cytokines. Methods: Patients received SBRT to 2-4 metastases and pembrolizumab ≤7 days after SBRT. Tumors <65cc received the full radiation dose (complete-Rx), whereas tumors > 65cc received partial-Rx. Landmark analysis was used to assess the relationship between tumor response and overall survival (OS). Multivariable analysis was performed for RS and circulating cytokines. Results: In the combined (expansion plus original) cohort, 97 patients (219 metastases) were analyzed and received SBRT+P. Forty-six (47%) patients received >1 partial-Rx treatment. There were 7 (7.2%) dose-limiting toxicities (DLTs). 1-year LF was 7.6% overall, and 13.3% and 5.4% for partial-Rx and complete-Rx tumors, respectively (hazard ratio 2.32, 95% confidence interval: 0.90-5.97, p = 0.08). The overall, unirradiated, and irradiated objective response rate was 22%, 12%, and 34%, respectively. Irradiated tumor response to SBRT+P was associated with prolonged OS; 1-year OS was 71% (responders), 42% (mixed-responders), and 0% (non-responders) (p < 0.01). High-RS was significantly associated with improved LF, progression-free survival (PFS), and OS. Elevated circulating IL-8 was independently associated with inferior PFS and OS. Conclusions: SBRT+P is safe in patients with large, advanced solid tumors. Additional studies are warranted to assess non-inferiority of complete versus partial irradiation of tumors in the setting of immunotherapy. Clinical trial information: NCT02608385.