The purpose of this study was to investigate the correlation between the seismocardiogram and cardiorespiratory fitness. Cardiorespiratory fitness can be estimated as VO2max using non-exercise algorithms, but the results can be inaccurate. Healthy subjects were recruited for this study. Seismocardiogram and electrocardiogram were recorded at rest. VO2max was measured during a maximal effort cycle ergometer test. Amplitudes and timing intervals were extracted from the seismocardiogram and used in combination with demographic data in a non-exercise prediction model for VO2max. 26 subjects were included, 17 females. Mean age: 38.3±9.1 years. The amplitude following the aortic valve closure derived from the seismocardiogram had a significant correlation of 0.80 (p<0.001) to VO2max. This feature combined with age, sex and BMI in the prediction model, yields a correlation to VO2max of 0.90 (p<0.001, 95% CI: 0.83–0.94) and a standard error of the estimate of 3.21 mL·kg−1·min−1 . The seismocardiogram carries information about the cardiorespiratory fitness. When comparing to other non-exercise models the proposed model performs better, even after cross validation. The model is limited when tracking changes in VO2max. The method could be used in the clinic for a more accurate estimation of VO2max compared to current non-exercise methods.