The objective of this work was to build a dataset of bulk milk fatty acids (FA) profile and related farming practices, with a wide territorial representativeness, in order to develop models to predict of milk FA at a farm scale. Data from three research centres were collected. A total of 1118 bulk milk samples coming from France, Northern Italy, Norway, Slovakia, and Slovenia were included in the dataset. For each milk sample, cows’ diet and herd characteristics were collected by on farm surveys. The wide variability in production conditions and animal feeding was also reflected by a great variation in milk FA profile. Total saturated FA (SFA) ranged from 52.1 to 79.2 g/100 g of FA, total monounsaturated FA (MUFA) from 18.0 to 42.3 g/100 g of FA, polyunsaturated FA (PUFA) from 1.8 to 7.5 g/100 g of FA. Also cis9trans11-CLA and C18:3n-3 had a wide variability, ranging from 0.1 to 2.8 and from 0.02 to 1.7 g/100 g of FA, respectively. The present work will allow to better understand which parameters could play in central role on milk FA composition at a farm scale.