Abstract Introduction: Synergistic drug combinations are widely sought as anti-cancer strategies because they yield higher therapeutic efficacy than additive or antagonistic combinations, enabling reduced doses. The effect of synergistic drugs (versus non-synergistic drugs) on the process of drug resistance evolution is not well understood, particularly for heterogeneous cell populations and for gradual phenotype transitions toward resistance. In this work, we use computational modeling of cancer cell numbers to interrogate the dynamics of resistance evolution during treatment with two-drug combinations. Methods: We developed computational models of cancer evolution under two-drug combination therapy, to study the development of drug resistance in cancers with varying heterogeneity or graded plasticity of resistance. A large population of cells was simulated to undergo infrequent but stochastic phenotype change over time, subject to a pharmacological dosing model of combination effects. Results: When synergistic and non-synergistic combinations were administered at equally effective doses in silico, drug resistance evolved more rapidly under the synergistic drugs. When we modeled graded plasticity by simulating multiple levels of drug resistance, results were similar except with additional bottlenecks before fully resistant cells could emerge. Meanwhile, as heterogeneity increased, synergism increased the speed of forming resistant subclones with higher fitness. Because synergistic treatments by nature exploit the simultaneous action of both drugs for their efficacy, the development of partial or single-drug resistance caused a disproportionate escape from drug effects and faster clonal expansion under synergistic therapy, compared with additive or non-synergistic therapies. Our simulations consistently showed that pairs of drugs with less synergism were able to prolong the time until double-resistance arose, provided the comparison combinations were dosed for equal initial efficacy. Conclusion: We identify dosing criteria whereby synergistic pairs of drugs actually have worse performance than additive or non-synergistic pairs, for delaying the onset of double-drug resistance. Dosing criteria are also shown for graded resistance and variable levels of population heterogeneity. Future work must examine non-heritable contributions to resistance, such as micro-environmental effects. Our theoretical work shows the need for further study of the divergence between short-term and long-term drug efficacy in the presence of clonal selection, and suggests increased attention to non-synergistic drug combinations as therapeutic candidates. Citation Format: Elysia C. Saputra, Lu Huang, Lisa Tucker-Kellogg. Synergistic drug combinations promote expansion of partially-resistant subpopulations in computational modeling of cancer heterogeneity and graded plasticity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1185.