Background: Real-world data have suggested that increases in overall survival (OS) for metastatic cancers over time have been driven by improvements in the groups of patients who already had the highest survival outcomes. To further investigate these identified trends, this research aimed to quantify how novel therapies for metastatic cancers impact disparities in OS outcomes. Methods: Novel therapies for metastatic breast (mBC), metastatic colorectal (mCRC), and metastatic non-small cell lung (mNSCLC) cancer, approved by the United States Food and Drug Administration (FDA) in the last 8 years based on data from phase III randomised controlled trials (RCTs), were identified. For each cancer type, 3 novel therapies were selected with inclusion based on the availability of published Kaplan-Meier data for OS, at least 100 patients in each arm of the corresponding RCT, and OS reaching less than 30% at the end of the RCT follow-up period. For each cancer type and for the novel therapies and comparator therapies separately, the published OS data were replicated from the Kaplan-Meier data and pooled using weights based on sample sizes to ensure each therapy contributed equally to the analysis. Parametric survival modelling was performed to address censoring. The disparity in OS was defined by the absolute difference between groups with the highest and lowest survival, with groups defined by stratifying OS in either 2 or 5 groups. The impact of the combined novel therapies was determined based on the change in OS disparity relative to the combined comparators. A probabilistic analysis was performed to quantify the impact of parameter uncertainty on the outcomes. Results: Novel therapies increased disparities in OS across the 3 cancers. For mBC, the disparities in OS increased by 5.4 (95%-confidence interval: 0.9; 9.9) months, which equates to a 23.0% increase from 24.2 to 29.6 months if OS was stratified in 2 groups, and by 13.9 (1.4; 26.6) months; 25.0% from 57.1 to 71.1 if OS was stratified in 5 groups. For mCRC, disparities increased by 2.8 (1.4; 4.2) months; 37.6% from 7.5 to 10.3 months and by 6.3 (2.9; 9.8) months; 38.9% from 16.5 to 22.8 months, for stratification based on 2 and 5 groups, respectively. For mNSCLC, disparities increased by 10.2 (3.3; 17.3) months; 33.6% from 31.0 to 41.3 months and by 19.5 (4.0; 35.5) months; 31.5% from 63.2 to 82.7 months, for stratification based on 2 and 5 groups, respectively. Conclusions: OS outcomes for recent FDA-approved drugs for metastatic cancers suggest that these novel therapies increase disparities in OS between the patient groups with the highest and lowest OS relative to their comparator therapies in RCTs. Although evidence in terms of these disparities is important, further research using (real-world) individual patient data is warranted to assess how, for example, socioeconomic variables affect the impact of therapies on health disparities among patient groups.