A total of 130 silages samples (53 of maize silages and 77 of grass silages ), which were ensiled with or without silage additives, with different soil contamination levels, with different weed percentages and with or without wilting, were used to evaluate the dry matter (DM) and crude protein (CP) ruminal degradability. The ruminal degradability of the samples was calculated from the corresponding in situ degradation parameters and from the measured passage rates of the silages fed to each experimental animal. The DM and CP degradation parameters were obtained using the logistic model of Van Milgen and Baumont. The fitting of the models to the kinetics of degradation and particle passage was carried out by non-linear regression. The value of the effective degradability, considering in the rumen simultaneously an outflow compartment and a mixing-reduction compartment, were calculated in both cases from an adaptation of the general procedure proposed by Ørskov and McDonald. A NIRSystems 6500 spectrometer was used for the prediction of the DM and the CP degradation characteristics of the samples. Calibration equations were obtained by modified partial least squares regression, using reflectance spectra transformed into the second derivative. The results showed that near infrared spectroscopy is a good method for predicting the DM and CP degradation characteristics. The calibrations for effective degradability of maize and grass silages indicated a high consistency.