Load modifying gait retraining strategies, such as changing the foot progression angle (FPA) to toe-in and toe-out gait, are used for people with medial knee osteoarthritis. The FPA can be measured using a pressure sensitive walkway (PSW), but inertial measurement units (IMUs) are considered more suitable for clinical use. This study evaluated the reliability and validity of an IMU system, to measure FPA under different gait retraining strategies. Twenty healthy participants walked a 10-m-long path using different gait strategies (natural (2), toe-out gait (1), toe-in gait (1)) during four 90-s trials. FPA was measured simultaneously with IMUs and a PSW, the latter considered the reference standard. There was good and excellent reliability for the IMUs and PSW FPA measurements, respectively (ICC: IMU, 0.89; PSW, 0.97). Minimal detectable change (MDC) was 4.5° for the IMUs and 2.7° for the PSW. Repeated measures ANOVA indicated a significant effect of gait type on FPA (p < 0.001), but not the measurement instrument (p = 0.875). Bland–Altman plots demonstrated the good agreement of both systems for the baseline condition, though the IMUs seemed to consistently overestimate the FPA value compared to the PSW. In conclusion, IMUs are a reliable and valid measurement system for measuring FPA under different gait retraining strategies. The differences between the systems are significant for all gait strategies, so the systems should not be used interchangeably.