Aim To investigate the association between the longitudinal changes of C-reactive protein-triglyceride-glucose index (CTI) and the longitudinal trajectories of relative fat mass (RFM) with the risk of new-onset stroke in middle-aged and elderly populations. Methods The study used data from the China Health and Retirement Longitudinal Study (CHARLS) for analysis. Based on two measurements of CTI in 2011 and 2015, the K-mean clustering was used to identify the categories of CTI changes. Based on three measurements of RFM in 1,3 and 2015, latent class mixed-growth modeling was applied to fit the RFM trajectory. Multivariate Logistic regression was used to assess the risk effects of CTI categories, RFM trajectories, and their combination on new-onset stroke. Relative excess risk analysis was conducted to evaluate the interaction between CTI categories and RFM trajectories. ROC curve analysis was performed to evaluate the discriminatory power of CTI categories and RFM trajectories for new-onset stroke. Results Of the 4 823 participants meeting the inclusion and exclusion criteria, 393 developed new-onset stroke during a median follow-up of 3 years. After adjusting for confounding factors, Logistic regression analysis showed that compared to category 1 with consistently low CTI levels, category 3 with consistently high CTI levels had a 70% increased risk of new-onset stroke (OR＝1.0,5%CI:1.24～2.34), while the RFM high stability group (trajectory 3) had a 62% increased risk compared to the low stability group (trajectory 1) (OR＝1.2,5%CI:1.04～2.64). Joint analysis showed that participants in category 3 and trajectory 3 had a significantly higher risk of new-onset stroke than those in category 1 and trajectory 1 (OR＝3.3,5%CI:1.98～6.96). The relative excess risk interaction analysis revealed that there was a significant additive interaction and synergistic effect between CTI categories and RFM trajectories. ROC curve analysis showed that CTI categories combined with RFM trajectories had a certain discriminatory power for new-onset stroke (AUC＝0.617). Conclusion The sustained high levels of CTI and RFM are significantly associated with an increased risk of new-onset stroke, and continuous monitoring of CTI and RFM is of great significance for the prevention and treatment of new-onset stroke.