Maximal oxygen uptake was determined in 62 boys each year from age 11 to 15 years. Multiple regression analyses were used to predict treadmill VO2 max at yearly age intervals from size measures and submaximal cycle heart rates. Derived equations were validated in an independent sample of 98 boys of the same age range. In boys aged 12 to 15 years approximately 80% or more of the variance in VO2 max was accounted for by age specific equations employing mass, height and skinfolds as explanatory variables. Prediction of VO2max was not as strong for 10 and 11 year olds, with less than 70% of the variance explained. Measures of VO2 or work rate and heart rate were not useful in predicting VO2max either by themselves or in conjunction with the size measures. The findings clearly demonstrated that submaximal exercise testing to predict VO2max, or cardiorespiratory fitness, in boys was of questionable value. The prediction of VO2max in boys aged 12 to 15 by in the independent sample was shown to have a mean absolute error of 6 to 8% while the value for age 11 y was 10%. The error is similar in magnitude to that of predicting VO2max with an exercise test in adults. The present age specific equations resulted in smaller errors of estimation than found in previous studies of VO2max estimation in children.