This study calculates efficiency scores for the Regional Innovation Systems (RIS) and Urban Innovation Systems (UIS) of Colombia using Data Envelopment Analysis (DEA) and identifies key determinants based on the Triple Helix model of innovation. We assess how regions and cities convert R&D investment and labor into measurable outputs such as patent activity and scientific publications. The results reveal significant heterogeneity in efficiency, with scores clustered at both extremes, suggesting that while some regions are highly efficient innovators, others lag considerably behind. Panel data models are used to explore explanatory factors behind this divergence. The findings indicate that regions with a broader supply of high-quality universities and denser business activity tend to perform better. In urban contexts, worker satisfaction emerges as a critical driver of innovation, highlighting the role of intrapreneurship—employees who innovate from within organizations—as a key mechanism. Additionally, the speed of legal proceedings enhances regional outcomes by reducing uncertainty and enabling a more agile and supportive environment for innovation-related activities. These results offer actionable insights for designing policies to strengthen Colombia’s innovation capacity at both the urban and regional levels.