Using ensemble ocean simulations, recent studies have shown that non-linear intrinsic oceanic processes are a source of chaotic intrinsic oceanic variability (CIOV). It was found that in eddy-active regions and at interannual timescales, this CIOV can be a significant fraction of total variability, and that as model resolution increases small-scale non-linearities can generate variability at large scales. The Eighteen Degree Water (EDW) is a mode water formed in the winter mixed layer within and south of the Gulf Stream. It is the most abundant T,S class of water in the surface North Atlantic and has been shown to be an important contributor to air-sea exchanges over the entire North Atlantic basin. Observational studies have shown that a significant part of the interannual variability of EDW cannot be explained by atmospheric variability. This motivates the present investigation of the importance of interannual CIOV in the total interannual EDW variability. The present study uses a NEMO-core, 1/4°, 50-member ensemble hindcast of the North Atlantic ocean with a realistic atmospheric forcing. This ensemble simulation is assessed using ARMOR3D, a 3-dimensional gridded observational product obtained using satellite altimetry and ARGO floats. In both datasets, the 3-dimensional structure of EDW is identified using physical criteria. This spatial structure is used to compute timeseries of the EDW’s total volume and average temperature, in each ensemble member and in the observational product. It is found that the ensemble simulation produces a realistic EDW, with a comparable total variability. In the ensemble simulation, the CIOV of integrated EDW properties is estimated from their time-averaged ensemble standard deviation, and is compared to the total variability estimated from the ensemble mean of the temporal standard deviations of all members. In the ensemble, CIOV accounts for 13% of the total interannual variability of EDW volume, and 44% of the total interannual variability of EDW temperature. Notably, this means that CIOV is a source of unquantifiable uncertainty in single-member ocean simulations. This suggests that a significant part of observed interannual variability may also be chaotic intrinsic in nature. This calls for a better parametrisation of chaotic variability in ocean simulations.