The feasibility of the three-dimensional (3D) cartilage regeneration technology based on the “steel (framework)-reinforced concrete (engineered cartilage gel, ECG)” concept has been verified in large animals using a decalcified bone matrix (DBM) as the framework. However, the instability of the source, large sample variation, and lack of control over the 3D shape of DBM have greatly hindered clinical translation of this technology. To optimize cartilage regeneration using the ECG–framework model, the current study explores the feasibility of replacing the DBM framework with a 3D-printed polycaprolactone (PCL) framework. The PCL framework showed good biocompatibility with ECG and achieved a high ECG loading efficiency, similar to that of the DBM framework. Furthermore, PCL-ECG constructs caused a milder inflammatory response in vivo than that induced by DBM-ECG constructs, which was further supported by an in vitro macrophage activation experiment. Notably, the PCL-ECG constructs successfully regenerated mature cartilage and essentially maintained their original shape throughout 8 weeks of subcutaneous implantation. Quantitative analysis revealed that the GAG and total collagen contents of the regenerated cartilage in the PCL-ECG group were significantly higher than those in the DBM-ECG group. The results indicated that the 3D-printed PCL framework—a clinically approved biomaterial with multiple advantages including customizable shape design, mechanical strength control, and standardized production—can serve as an excellent framework for supporting the 3D cartilage regeneration of ECG. This provides a feasible novel strategy for the clinical translation of ECG-based 3D cartilage regeneration.