For many years we are witnessing a lively debate on the existence and extent of convective overshooting, mainly in the cores of main-sequence stars. This is an important issue, since even a small amount of overshooting increases considerably the mass of the finally hydrogen exhausted core and lenghthens the main-sequence lifetime correspondingly. The available evolutionary calculations assume either moderate overshooting, d/Hp = 0.25, (d = overshooting distance, Hp = pressure scale height; Maeder & Meynet 1988) or strong overshooting, d/Hp ≈ 0.50 (Bertelli et al. 1986). Presently theory is unable to quantify the exact amount of overshooting, and one has to resort to empirical determinations.

Recently, Stothers (1991) collected all available information from the literature on stellar parameters and evolutionary calculations and concluded that, within the errors, d/Hp = 0 is an acceptable result, with a conservative upper limit of d/Hp < 0.2. However, such an approach is hampered by observational errors (like distance or temperature uncertainties, rotation) that are difficult to quantify and that may mask any definitive result. Detailed investigations of detached binaries may help in this matter (Andersen et al. 1990) but the number of suitable binary systems is probably not very large.