Protein digestibility, a key indicator of dietary protein quality for human nutrition, can be estimated using an in vitro digestion model, however its definition and determination remain variable across studies. The present study aimed to determine the contribution of the endogenous nitrogen (N) to the plant and animal protein digestibility values obtained in vitro. 15 N-labelled gluten and caseins (4, 8 and 16 % of the model meal) were used to differentiate dietary and endogenous N and were digested using the INFOGEST in vitro digestion model with no oral phase. The dietary and endogenous N were measured before and during digestion after centrifugation and 10 kDa ultrafiltration. The proteolysis degree was measured by the OPA method. The endogenous and dietary N were determined by elemental analyser coupled with isotopic ratio mass spectrometry. Apparent and true digestibility were determined and values of 135, 92 and 71 % for apparent vs. 78, 69, 60 % for true digestibility were obtained for 4, 8 and 16 % dietary protein level, respectively, with a significant effect of protein level. Differences between apparent and true digestibility pointed out the important contribution of the endogenous nitrogen. Our results showed that 40 % of the N below 10 kDa, i.e., the digestible fraction, were from endogenous origin (i.e. from the pancreatin) and was even present before digestion. An average value of 27 % for pancreatin N autolysis was estimated independently of the protein levels or sources. The use of 15 N-labelled protein to evaluate in vitro protein digestibility highlighted the important contribution of the endogenous N, in particular when low dietary protein solution (4 %) are digested. This gives new keys to overcome drawbacks of in vitro models for determining protein digestibility.