Background: It remains unknown if flywheel assisted squats can be reliably utilised to increase power outputs and if such outputs are related. Objectives: Compare assisted and unassisted flywheel squat peak power outputs, determine their reliability and analyse the relationship of the delta difference between peak power outputs during the squats. Methods: Twenty male athletes attended the laboratory six times – performing three sets of eight repetitions of assisted and unassisted squats during 2 familiarisation sessions and then three sets of eight repetitions during experimental sessions 3-6 (2 sessions for unassisted and assisted squat in randomised order, respectively). Results: Concentric and eccentric peak power were significantly greater during assisted squats (both p < 0.001, d = 1.59, d = 1.57, respectively). Rate of perceived exertion (RPE) (p = 0.23) and eccentric:concentric (E:C) ratio (p = 0.094) did not differ between squat conditions. Peak power measures obtained excellent reliability, while RPE and E:C ratio estimates were rated as acceptable to good, with greater uncertainty. A large to very large correlation (r = 0.77) was found between concentric and eccentric peak power delta difference of assisted and unassisted squats. Conclusions: Greater concentric outputs during assisted squats induce greater eccentric outputs and obtain greater mechanical load. Peak power is a reliable metric for monitoring flywheel training, whereas the E:C ratio should be used with caution. Eccentric and concentric peak power are strongly related during flywheel squats, evidencing the need to maximise the concentric output to enhance the eccentric output.