Purpose: To compare methods of monitoring and prescribing on-water exercise intensity (heart rate [HR], stroke rate [SR], and power output [PO]) during sprint kayak training. Methods: Twelve well-trained flat-water sprint kayak athletes completed a preliminary on-water 7 × 4-min graded exercise test and a 1000-m time trial to delineate individual training zones for PO, HR, and SR into a 5-zone model (T1–T5). Subsequently, athletes completed 2 repeated trials of an on-water training session, where intensity was prescribed based on individual PO zones. Times quantified for T1–T5 during the training session were then compared between PO, HR, and SR. Results: Total time spent in T1 was higher for HR (P < .01) compared with PO. Time spent in T2 was lower for HR (P < .001) and SR (P < .001) compared with PO. Time spent in T3 was not different between PO, SR, and HR (P > .05). Time spent in T4 was higher for HR (P < .001) and SR (P < .001) compared with PO. Time spent in T5 was higher for SR (P = .03) compared with PO. Differences were found between the prescribed and actual time spent in T1–T5 when using PO (P < .001). Conclusions: The measures of HR and SR misrepresented time quantified for T1–T5 as prescribed by PO. The stochastic nature of PO during on-water training may explain the discrepancies between prescribed and actual time quantified for power across these zones. For optimized prescription and monitoring of athlete training loads, coaches should consider the discrepancies between different measures of intensity and how they may influence intensity distribution.