Ruminal nitrogen degradation and intestinal digestibility (ID) of the undegraded nitrogen of three sunflower meals were determined on three wethers fitted with rumen cannulae and T-type duodenal cannulae using nylon bags. Meals were obtained from semi-dehulled seeds by conventional hexane extraction (samples SD1 and SD2) or from whole seeds by a discontinuous procedure of pressing and hexane extraction (sample W), which causes a superior thermal effect. Therefore, effective degradability of nitrogen for the W sample (0.537) was lower (P < 0.001) than for conventional meals. Between the latter, SD2 had a lower value (P = 0.019) than SD1 (0.776 and 0.812, respectively). ID decreased in all meals (P < 0.001) as the ruminal incubation time (t) increased. This evolution could be described accurately by an exponential curve as ID = s + he(-kit). A method is proposed for estimating the proportion of undegraded ruminal nitrogen digested in the intestines (Di) from 1) the above equation, 2) the undegradable (r) and the insoluble and potentially degradable (b) nitrogen contents of the feed and the degradation rate of the last fraction (k(d)), and 3) the rumen outflow rate of particles (k(p)). The Di value is shown to be: [equation: see text] The percentages of nitrogen from digested feed in the intestines obtained with this method were 15.1, 17.2 and 39.0 for SD1, SD2 and W, respectively. Resulting effective ID values of undegraded nitrogen were 0.804, 0.767 and 0.844. Undigested nitrogen after ruminal and intestinal incubations decreased in linear and quadratic form in all meals as ruminal incubation time increased.