Heart rate samples (n = 4500-8000) from wearables were compared to electrocardiography during a steady-state ruck (Ruck-S), maximal effort ruck (Ruck-M), submaximal cycle (Cycle), and Tabata Circuit. One device was worn at each location (wrist: Polar Grit-X, Garmin Fenix 6; chest-straps: Polar H10, Garmin HRM-Pro; armband: Polar Verity). Comparisons were made via percent error (MAPE) [less than or equal to]5%, Concordance Correlation Coefficient (CCC), and ordinary least product regressions (OLP). H10 demonstrated strong agreement for all movements (MAPE = 1.28-3.40%, CCC = 0.93-0.99). During Ruck-S, Ruck-M, and Cycle, HRM-Pro (MAPE = 1.96-3.73%, CCC = 0.95-0.99) and Verity (MAPE = 1.84-5.36%, CCC = 0.98-0.99) demonstrated strong agreement. Fenix-6 demonstrated low MAPE (4.23-5.44%) and moderate to strong CCC (0.76-0.96) for Ruck-S, Ruck-M, and Cycle, while Grit-X had poor agreement (MAPE = 8.49-16.45%; CCC = 0.24-0.78). Tabata Circuit had the worst disagreement for all devices. Overall, chest straps and armbands demonstrated the strongest agreement, and should be worn when precise heart rate training is necessary.