New feeding systems for ruminants are based on the estimation of the quantity of protein that can be absorbed by the small intestine. This protein has two origins: dietary or microbial. The measurement of food protein which escapes rumen degradation is probably the major element in the evaluation of food protein value. The in situ incubation in nylon bags (Ørskov and McDonald, 1979) is the technique widely utilized to determine rate and extent of protein degradation. However, the technique is difficult to manage as it requires cannulated animals and can furnish reproducible data only utilizing standardized procedures. Although it is possible to standardize the amount and particle size of food sample, bag characteristics, incubation times and calculation model, it is, on the other hand, very difficult to standardize the basal diet (in our opinion to evaluate the diet only in function of forage/concentrate ratio and crude protein content is inadvisable). Thus for many years various alternative methods which are easier to manage and to standardize (chemical, enzymatic and microbiological procedures) have been developed (Nocek, 1988), to replace the in situ method. The most commonly used enzyme is an endoprotease from Streptomyces griseus (Krishnamoorthy et al, 1983; Poos-Floyd et al., 1985; Aufrère et al, 1991, Infascelli et al, 1993). However, some of the authors (Infascelli et al., 1993 plus Poos-Floyd et al., 1985) agree that this bacterial protease is of little consequence when predicting protein food degradation due to its unspecificity. In addition Poos-Floyd et al., 1985) suggest discarding the endoprotease from Streptomyces griseus due to the variability of results. On the contrary, the latter authors, evaluating rumen protein escape and proportion of degraded nitrogen (N) with enzymatic methods contemporarily, found ficin, a tiolic enzyme, preferable. Recently, Tomànkovà and Kopêcný (1995) have developed a method using another tiolic enzyme, bromelain, to predict rumen protein degradability. Kosmala et al. (1996) in agreement with Poos-Floyd et al. (1985) indicate ficin as the protease to estimate rumen degradability. Because these enzymes allow measurement of crude protein solubility at fixed incubation times, we think that the latter should not be necessarily correlated with the effective degradability that originates from combining degradability kinetics with rumen outflow rate. Thus, the aim of the present work was to compare the bromelain and ficin methods for estimating N solubility, particularly with regard to the variability of results.