Eclipsing binary stars provide highly accurate measurements of the fundamental physical properties of stars. They therefore serve as stringent tests of the predictions of evolutionary models upon which most stellar age determinations are based. Models generally perform very well in predicting coeval ages for eclipsing binaries with main-sequence components more massive than ≈1.2 M⊙; relative ages are good to ~5% or better in this mass regime. Low-mass main-sequence stars (M < 0.8 M⊙) reveal large discrepancies in the model predicted ages, primarily due to magnetic activity in the observed stars that appears to inhibit convection and likely causes the radii to be 10–20% larger than predicted. In mass-radius diagrams these stars thus appear 50–90% older or younger than they really are. Aside from these activity-related effects, low-mass pre–main-sequence stars at ages ~1 Myr can also show non-coevality of ~30% due to star formation effects, however these effects are largely erased after ~10 Myr.