Innovation imbued in every sector in every part of the world is essential to growth and development. The National Innovation Systems (NIS) use their resources to support economies in different countries foster a novel environment. Therefore, this study is an attempt to examine the efficiency of NIS as measured by scholars in the past using Data Envelopment Analysis (DEA). Through a systematic literature review, this study aims to show the current dearth of studies on the efficiency of NIS. The literature is categorized accordingly and provides a framework with recommendations for future research. With the advent of technical tools, DEA programming methods other than traditional DEA such as network, relational network, dynamic network, slack based model, and super efficiency DEA have emerged. This enables the calculation of innovation efficiency to be efficient and reliable. As a result, DEA is regarded as a powerful tool for assessing the relative efficiency of NIS, which employs multiple inputs to generate multiple outputs. The results also suggest that future research is needed on the efficiency of innovation by paying more attention to cross-countries studies based on regions, geographical areas, participation in free trade blocs, and a group of tie-up countries involved, especially with comparative analyses regardless of the country classification. Another important observation is that there are few studies that focus on the efficiency of middle- and low-income countries. The comparative analysis of innovation efficiency across income groups enables NIS to benchmark itself against best-in-class innovators and improve their innovation performance and ranking. These findings provide an opportunity to further investigate how NIS add value and sustainability to countries by improving resource management capabilities to improve innovation efficiency.