Objective: The EPAZ study (NCT01861951) showed recently that pazopanib was non-inferior to doxorubicin in patients ≥60 years treated in first line for advanced soft tissue sarcoma . The current post-hoc analysis aimed to assess the prognostic impact of frailty. Methods: Geriatric assessments were evaluated at baseline. Age >75 years, liposarcoma, ECOG = 2, G8 ≤14, instrumental activities of daily living (IADL) ≥1 and Charlson Comorbidity Index ≥2 were tested for their impact on progression-free survival (PFS), overall survival (OS), CTCAE grade 3/4 adverse events (AEs) or serious AEs (SAEs), using univariate and multivariate analysis models. Results: univariate analysis showed an increased risk of grade 3/4 AEs and SAEs for ECOG = 2, G8 score ≤14 or IADL ≥1, independent of treatment. The multivariate analysis exhibited for pazopanib a significantly reduced risk for grade 3/4 AEs (HR 0.53; p = 0.033), and in patients with G8 ≤14 an increased risk for SAEs (HR 2.67; p = 0.011). In the multivariate analysis, G8 ≤14 was a negative prognostic factor for PFS (HR 1.82; p = 0.009) and IADL ≥1 for OS (HR 2.02; p = 0.007). ECOG = 2 was the strongest negative predictor for PFS (HR 4.39; p = 0.001) and OS (HR 3.74; p = 0.004). Neither age nor Charlson Comorbidity Index showed any impact on PFS, OS, incidence of grade 3/4 AEs or SAEs. Conclusions: This post hoc analysis demonstrated that age is not a denominator for outcome or toxicity in elderly patients with soft tissue sarcoma . Instead, geriatric and functional assessments should be used to counsel patients and tailor therapy to individual needs. Moreover, pazopanib has a reduced risk for grade 3/4 AEs compared to doxorubicin. © 2022 Elsevier Ltd