Not yet reviewed by Pith; the record is open.

Eclipsing binaries with M-type components are still rare objects. Strong observational biases have made that today only a few eclipsing binaries with component masses below 0.6 Msun and well-determined fundamental properties are known. However, even in these small numbers the detailed comparison of the observed masses and radii with theoretical predictions has revealed large disagreements. Current models seem to predict radii of stars in the 0.4-0.8 Msun range to be some 5-15% smaller than observed. Given the high accuracy of the empirical measurements (a few percent in both mass and radius), these differences are highly significant. I review all the observational evidence on the properties of M-type stars and discuss a possible scenario based on stellar activity to explain the observed discrepancies.