Inhalation and ingestion of ¹³⁷Cs and other long-lived radionuclides can occur after large-scale accidental or malicious radioactive contamination incidents, resulting in a complex temporal pattern of radiation dose/dose rate, influenced by radionuclide pharmacokinetics and chemical properties. High-throughput radiation biodosimetry techniques for such internal exposure are needed to assess potential risks of short-term toxicity and delayed effects (e.g., carcinogenesis) for exposed individuals. Previously, we used γ-H2AX to reconstruct injected ¹³⁷Cs activity in experimentally-exposed mice, and converted activity values into radiation doses based on time since injection and ¹³⁷Cs-elimination kinetics. In the current study, we sought to assess the feasibility and possible advantages of combining γ-H2AX with transcriptomics to improve ¹³⁷Cs activity reconstructions. We selected five genes (Atf5, Hist2h2aa2, Olfr358, Psrc1, Hist2h2ac) with strong statistically-significant Spearman's correlations with injected activity and stable expression over time after ¹³⁷Cs injection. The geometric mean of log-transformed signals of these five genes, combined with γ-H2AX fluorescence, were used as predictors in a nonlinear model for reconstructing injected ¹³⁷Cs activity. The coefficient of determination (R²) comparing actual and reconstructed activities was 0.91 and root mean squared error (RMSE) was 0.95 MBq. These metrics remained stable when the model was fitted to a randomly-selected half of the data and tested on the other half, repeated 100 times. Model performance was significantly better when compared to our previous analysis using γ-H2AX alone, and when compared to an analysis where genes are used without γ-H2AX, suggesting that integrating γ-H2AX with gene expression provides an important advantage. Our findings show a proof of principle that integration of radiation-responsive biomarkers from different fields is promising for radiation biodosimetry of internal emitters.