Age-based prediction equations of maximal heart rate (HR), such as the popular formulas Fox's 220-age, or Tanaka's 208-0.7 × age, have been widely used in various populations. Surprisingly, so far these equations have not been validated in marathon runners, despite the importance of the role of HR for training purposes in endurance running. The aim of the present study was to examine the validity of Fox and Tanaka equations in a large sample of women and men recreational marathon runners. Participants ( = 180, age 43.2 ± 8.5 years, VO 46.8 mL/min/kg, finishers in at least one marathon during the last year) performed a graded exercise test on a treadmill, where HR was measured. Measured HR correlated largely with age in the total sample ( = -0.50, < 0.001), women ( = -0.60, < 0.001) and men ( = -0.53, < 0.001). In women, a large main effect of method on HR ( = 0.001, η = 0.294) was shown with measured HR lower than Fox-HR (-4.8 bpm; -8.4, -1.3) and Tanaka-HR (-4.9 bpm; -8.1, -1.8). In men, a moderate effect of assessment method on HR was found ( = 0.001, η = 0.066) with measured HR higher than Fox-HR (+2.8; 1.0, 4.6), Tanaka-HR higher than Fox-HR (+1.2; 0.7, 1.7). Based on these findings, it was concluded that Fox and Tanaka' formulas overestimated HR by ~5 bpm in women, whereas Fox underestimated HR in men by ~3 bpm. Thus, we recommend the further use of Tanaka's formula in men marathon runners. In addition, exercise physiologists and sport scientists should consider the observed differences among various assessment methods when performing exercise testing or prescribing training program relying on HR.