Previous studies have demonstrated that during lower-body exercise the percentage of heart rate reserve (%HRR) is equivalent to the percentage of the oxygen consumption reserve (% V˙ O_2R) but not to a percentage of the peak oxygen consumption (% V˙ O_2peak). The current study examined these relationships in trained surfboard riders (surfers) during upper-body exercise. Thirteen well-trained competitive surfers performed a stepwise, incremental, prone arm-paddling exercise test to exhaustion. For each subject, data obtained at the end of each stage (i.e., HR and V˙ O₂ values) were expressed as a percentage of HRR, V˙ O_2peak, and V˙ O_2R respectively and used to determine the individual %HRR-% V˙ O_2peak and %HRR-% V˙ O_2R relationships. Mean slope and intercept were calculated and compared with the line of identity (slope=1, intercept=0). The %HRR versus % V˙ O_2R regression mean slope (0.88±0.06) and intercept (20.82±4.57) were significantly different ( p <0.05) from 1 and 0, respectively. Similarly, the regression of %HRR versus % V˙ O_2peak resulted in a line that differed in the slope ( p <0.05) but not in the intercept ( p =0.94) from the line of identity. Predicted values of %HRR were significantly higher ( p <0.05) from indicated values of % V˙ O_2R for all the intensities ranging from 35% to 95% V˙ O_2R. Unlike results found for lower-body exercise, a given %HRR during prone upper-body exercise was not equivalent to its corresponding % V˙ O_2R. Thus, to ensure more targeted exercise intensity during arm-paddling exercise, individual HR- V˙ O₂ equations should be used.