After maintenance, turbofan engines are subjected to a performance acceptance test in an indoor test-cell to demonstrate that corrected performance thresholds are met. The same indicators are also monitored after on-wing installation. Despite corrections for operation conditions, differences are observed between test-cell and subsequent on-wing performance. A comprehensive list of potential root causes for those differences was investigated using data-driven analyzes, theory and simulations. The main root causes are thermal effects, resulting from the lack of thermal stabilization during on-wing operation, and seal run-in, resulting from the initial decrease of effectiveness of replaced seals. Aircraft sensor bias and test-cell correction factors are expected to also contribute considerably. Engine bleed air and power extraction effects are negligible. The impact of inaccurate or missing throttle, temperature and humidity corrections was eliminated by application of proposed engine-specific customized corrections, which served as a successful proof of concept for improved on-wing monitoring accuracy.