The estimation of maximal oxygen consumption (max Vo2) from a simple submaximal test has been of interest for many years, especially for middle-aged men. The object of the present study was to compare the prediction of max Vo2obtained by simple regression with that obtained by stepwise multiple regression. The subjects, 13 middle-aged men, were exercised on a bicycle ergometer to max Vo2(P < 0.10) as determined by the Haldane-Douglas bag method. Heart rate, systolic and diastolic blood pressure, expired volume, expired CO2and O2were determined. The multiple regression equation predicted max Vo2better than the simple regression equation because several of the cardiovascular and respiratory variables are significant predictors and do not contain identical information. For middle-aged men, the multiple regression equation provided a correlation with max Vo2that was significantly superior to that obtained by simple regression. This suggests that fallibility may be reduced in the prediction of max Vo2by relying on more than a single cardiovascular or respiratory variable in the prediction equation.