Objective To investigate body temperature trajectories and identify predictors during the early phase (first 8 h) of continuous renal replacement therapy (CRRT) in critically ill patients, and to provide a foundation for developing personalized temperature management protocols. Methods A retrospective study was conducted involving 263 critically ill patients who received CRRT in the ICU between January 2022 and October 2024. Body temperature data were collected 1 hour before CRRT initiation and at intervals during the first 8 hours of therapy. Latent growth mixture modeling (LGMM) was used to identify distinct body temperature trajectories.Multinomial logistic regression was then employed to determine predictive factors for trajectories. Results Three trajectories were identified:stable (80.2%), progressive decline (11.8%), and fluctuating (8.0%). The mean arterial pressure (MAP) and baseline body temperature were predictors of the trajectory patterns (all P<0.05). Conclusion Body tempe-rature trajectories during the initial 8 hours of CRRT exhibit significant heterogeneity among critically ill patients. Healthcare providers can utilize patients′ baseline MAP and body temperature to predict their likely thermal trajectory. This enables the implementation of tiered, proactive temperature management strategies aimed at maintaining thermal stability.