Purpose: To compare the intensity and physiological responses of deoxygenated hemoglobin breaking point ([HHb]BP) and anaerobic threshold (AnT) during an incremental test and to verify their association with 2000-m rowing-ergometer performance in well-trained rowers. Methods: A total of 13 male rowers (mean [SD] age = 24 [11] y and = 63.7 [6.1] mL·kg−1·min−1) performed a step incremental test. Gas exchange, vastus lateralis [HHb], and blood lactate concentration were measured. Power output, , and heart rate of [HHb]BP and AnT were determined and compared with each other. A 2000-m test was performed in another visit. Results: No differences were found between [HHb]BP and AnT in the power output (236 [31] vs 234 [31] W; Δ = 0.7%), 95% confidence interval [CI] 6.7%), (4.2 [0.5] vs 4.3 [0.4] L·min−1; Δ = −0.8%, 95% CI 4.0%), or heart rate (180 [16] vs 182 [12] beats·min−1; Δ = −1.6%, 95% CI 2.1%); however, there was high typical error of estimate (TEE) and wide 95% limits of agreement (LoA) for power output (TEE 10.7%, LoA 54.1–50.6 W), (TEE 5.9%, LoA −0.57 to 0.63 L·min−1), and heart rate (TEE 2.4%, LoA −9.6 to 14.7 beats·min−1). Significant correlations were observed between [HHb]BP (r = .70) and AnT (r = .89) with 2000-m mean power. Conclusions: These results demonstrate a breaking point in [HHb] of the vastus lateralis muscle during the incremental test that is capable of distinguishing rowers with different performance levels. However, the high random error would compromise the use of [HHb]BP for training and testing in rowing.