The purpose of this study was to assess the validity of predicting maximal oxygen uptake(VO(2max)) from sub-maximal VO(2) values elicited during a perceptually-regulated exercise test. We hypothesised that the strong relationship between the ratings of perceived exertion (RPE) and VO(2) would enable VO(2max) to be predicted and that this would improve with practice. Ten male volunteers performed a graded exercise test (GXT) to establish VO(2max) followed by three sub-maximal RPE production protocols on a cycle ergometer, each separated by a period of 48 h. The perceptually-regulated trials were conducted at intensities of 9, 11, 13, 15 and 17 on the RPE scale, in that order. VO(2) and HR were measured continuously and recorded at the end of each 4 min stage. Individual's RPE values yielded correlations in the range 0.92-0.99 across the three production trials. There were no significant differences between measured VO(2max) (48.8 ml.kg(-1).min(-1)) and predicted VO(2) max values (47.3, 48.6 and 49.9 ml.kg(-1).min(-1), for trials 1, 2 and 3, respectively) when VO(2) max was predicted from RPE values of 9-17. The same was observed when VO(2max) was predicted using RPE 9-15. Limits of agreement (LoA) analysis on actual and predicted VO(2max) values (from RPE 9-17) were (bias+/-1.96xSDdiff) 1.5+/-7.3, 0.2+/-4.9 and -1.2+/-5.8 ml.kg(-1).min(-1), for trials 1, 2 and 3, respectively. Corresponding LoA values for actual and predicted VO(2max) (from RPE 9-15) were 5.4+/-11.3, 4.4+/-8.7 and 2.3+/-8.4 ml.kg(-1).min(-1), respectively. The data suggest that a sub-maximal, perceptually-guided, graded exercise protocol can provide acceptable estimates of maximal aerobic power, which are further improved with practice in fit young males.