There has been much debate in the recent scientific literature regarding the possible ability to increase gross efficiency in cycling via training. Using cross-sectional study designs, researchers have demonstrated no significant differences in gross efficiency between trained and untrained cyclists. Reviewing this literature provides evidence to suggest that methodological inadequacies may have played a crucial role in the conclusions drawn from the majority of these studies. We present an overview of these studies and their relative shortcomings and conclude that in well-controlled and rigorously designed studies, training has a positive influence upon gross efficiency. Putative mechanisms for the increase in gross efficiency as a result of training include, muscle fibre type transformation, changes to muscle fibre shortening velocities and changes within the mitochondria. However, the specific mechanisms by which training improves gross efficiency and their impact on cycling performance remain to be determined.