The present study re-evaluated the accuracy of standards for maximal exercise capacity (Wcap) recently reported from our laboratory by examining the interaction between height and age on Wcap achieved and predicted in 1,071 subjects (732 males and 339 females). They underwent an incremental exercise test on a cycle ergometer using the same incremental protocol and exercise mode as the previous study, and were retrospectively judged to be normal. Although Wcap predicted was either not significantly different (males) or underestimated Wcap by less than 5% (females, p less than 0.05), significant differences were found in subjects at the extremes of the population ranges for height and age. The influences of age 9yr) and height (m) were found to be nonlinear and interactive, as described by the equations: Males: Wcap = 1506*Ht2.70*Age-0.46(r = 0.78) (lower limit 81% pred) Females: Wcap = 969*Ht2.80*Age-0.43(r = 0.77) (lower limit 79% pred) Wcap (kpm/min) predicted by these equations was compared to Wcap achieved by the 100 subjects who took part in the original study; no significant differences were found (paired t test, p less than 0.05). The interactive influences of age and height expressed by the equations are more plausible from a biological point of view than the linear, additive relationships previously described. The equations should be more reliable than previous equations for patients referred for exercise testing.