The Mimas-Thethys and Enceladus-Dione Systems A. Vienne, L. Duriez Laboratoire d'Astronomie. Universite des Sciences et Techniques de Lille. F59000 Lille We have obtained in Lille a theory of the motion of the eight major satellite s of Saturn with the precision of a few kilometers over one century within the c onsidered physical model. This theory has been adjusted to the Earth based obser vations of these satellites. We have found in the mean longitude of Mimas some very long period terms (unt il 700 years). The argument of these terms contains the pericenter of Tethys, th en the amplitude of such terms depends on its eccentricity. These terms were mor e numerous in our previous works because the eccentricity of Tethys was 5 times greater than that found now. But, these long periods correspond to small divisor s, so they exist in any case even if the eccentricity of Tethys is very small. The solution of the system Enceladus-Dione appears much simpler, as we have f ound no small divisor. In this paper, we study the sensitivity of the solution of these four satelli tes, not only to variations of the initial conditions but also to variations of the flateness of Saturn. The results of preliminary computations is presented. A discussion of these results is done in relation with the problem of the origin of the resonance.