We present the results of the classification procedure described in Piquard's PhD thesis applied to the 1091 stars presented in Piquard et al. (2001. Cat. <J/A+A/373/576>). In those six tables we give indication of variability type for all the stars detected as variables in Piquard et al. (2001, Cat. <J/A+A/373/576>). First of all, we looked for periodic signals in the light curves using adapted Renson's (Renson, 1978A&A....63..125R) and Stellingwerf's (Stellingwerf, 1978ApJ...224..953S) methods applied to the tree different photometric bands from Tycho (Cat. <I/239>): T, B_T_, V_T_ (the T band is defined from the added count-rates in the B_T_ and V_T_ channels, Grossmann et al.. 1995A&A...304..110G). Then we developed a semi-automatic method using 5 parameters (Period, color index, reduced proper motion from Tycho-2 catalogue (Cat. <I/259>), 2 indicators of the shape of the light curve) and a maximum likelihood method, combined with a careful look on the light curves. Finaly, the identification of the variability type is kept only if the resulting light curve is convincing (periodic the star has the P status, if not periodic the star has the U status); else we need more information about the star and is has the A status. All this identifications are indications since the quality of the light curves are often poor, particularly when the star is fainter than T=10.