Objective: The number of early-onset colorectal cancer survivors (EOCRCs) is increasing. The primary aim of rehabilitation after battling cancer is to enable patients to return to work, as they constitute a significant contributor to societal productivity. A predictive model was developed to identify priority populations requiring intervention and refine responses to increase their capacity to return to work after undergoing treatment for EOCRC. Methods: The baseline information was collected before patients were discharged at the end of their treatment course. The data of patients who returned to work were collected at 1 and 2 years after discharge. A predictive variable model was developed via binary logistic regression. The TRIPOD checklist was used. Results: At 1 year, 64.7% of the EOCRC survivors had returned to work. Male sex, education, return to work self efficacy, re-entry readiness and social support increased the possibility of returning to work; higher levels of self-perceived fatigue and lower levels of family care decreased the possibility of returning to work within the 1-year model. At 2 years, 72.8% of the EOCRC survivors had returned to work. In the 2-year model, education, self-transcendence, return to work self efficacy, re-entry readiness and occupational environment increased the possibility of returning to work; self-perceived fatigue and psychosocial adjustment decreased the possibility of returning to work. Conclusions: The results of this study can guide early assessment and intervention for EOCRC survivors, to facilitate their return to work.