To evaluate and discuss the accuracy of submaximal exercise-based equations to predict maximum oxygen uptake (V˙o2max), validated using direct gas analysis, in older apparently healthy adults. Studies were identified by searching 5 electronic databases and manually scanning reference lists of included articles from the respective inception of each database through April 2015. Studies were included if they used at least 1 submaximal exercise-based variable in the prediction, the actual V˙o2max was directly measured using a gas analysis device, and if participants were apparently healthy older adults (mean age ≥65y). Eligible studies were required to report at least 1 validity statistic (eg, Pearson product-moment correlation [r]) and either a predicted and measured V˙o2max value or a directional significant difference between the measured and predicted V˙o2max values. No limits were placed on year of publication, but only full-text, published articles in the English language were included. Nine articles and 13 equations were retained from the systematic search strategy. If the same prediction equation was used across multiple trials, data from the most accurate trial were reported. Submaximal equations predicted directly measured V˙o2max with a moderate to strong correlation strength (r range, 0.4-0.9). Predicted V˙o2max significantly differed from directly measured in 2 of the 13 equations. The preferred mode of ergometry was walking or running (7 equations); a stepping protocol was the most accurate (R(2)=0.9, not significant between predicted and measured V˙o2max). Factors to consider when choosing a submaximal exercise-based equation are the accuracy of the equation, the population tested, the mode of ergometry, the equipment availability, and the time needed to conduct familiarization sessions.