Measuring the minimum of an eclipse light curve has practical and theoretical difficulties. The two common methods - Kwee & van Woerden and polynomial fitting, often disagree significantly. This talk will discuss various methods, their good and weak points, and when and how best used.