Current methods of determining feed efficiency in ruminants are laborious and difficult to measure and consequently, alternative biomarkers are being explored. Based on the idea of isotopic fractionation, we measured the natural abundance of δ15N and δ13C (‰) in plasma proteins of 54 Charolais cattle and performed a regression analysis against different feed efficiency indices to determine their potential a as biomarker to predict between-animal variations of feed efficiency. The cattle were examined for feed conversion efficiency (FCE) and residual feed intake (RFI), beginning at 11-13 months of age, over two years (n=20 in 2014 and n=34 in 2015). Despite identical dietary constituents, the crude protein composition varied across the two years (13 vs 15% DM in 2014 and 2015, respectively) and consequently, animals from each year were analyzed separately. The natural abundance of δ15N in plasma proteins was higher in 2014 cattle (δ15N av.=6.22), vs 2015 (δ15N av.=5.54). Whereas, δ13C was higher in animals from 2015 (δ13C av.=-24.61) vs 2014 (δ13C av.=- 25.13). A significant negative correlation was observed between δ15N and FCE in animals from both 2014 and 2015 (R2=0.62 and R2=0.52, respectively). Similarly, δ13C in plasma proteins showed a moderate negative correlation with FCE in 2014 (R2=0.34), but no correlation with FCE was observed in 2015 (R2=0.09). Nor was a correlation observed between RFI and either δ15N (R2=0.07 and R2=0.09) or δ13C (R2=0.04 and R2=0.005) in 2014 or 2015. The repeatability of the relationship between δ15N in plasma protein and feed conversion efficiency in the two groups of cattle indicates its potential as a biomarker between-animal variations of feed efficiency, measured as FCE but not RFI, in ruminants.