Four cows were used in a 4 ] 4 Latin square design with a 2 ] 2 factorial arrangement of treatments to examine the e †ects of the extent of silage fermentation and post-ruminal casein infusion on the arterial concentrations and mammary arterialÈvenous (AÈV) di †erence of energy-yielding nutrients and plasma amino acids. Of the two silages one was preserved with formic acid (FAsilage) and the other with an inoculant of lactic acid bacteria (LAB-silage). Both silages were given ad libitum with 8 kg day~1 of barley, without or with duodenal casein infusing (400 g day~1). Blood was sampled from the coccygeal vessel and milk vein. FA-silage less fermented than LAB-silage as indicated by higher concentration (g kg~1 dry matter) of water soluble carbohydrates (178 vs 68) and lower concentration of lactic acid (32 vs 147). Plasma concentration of glucose was lower (P \ 0É05) and concentrations of essential amino acids (P \ 0É001) and blood b-hydroxybutyrate (BHB) (P \ 0É01) were higher with FA-silage than with LAB-silage. There was a strong linear relationship between arterial concentration and mammary AÈV di †erence of blood metabolites, except BHB, and amino acids. The relationship between arterial concentration and mammary AÈV di †erence of glucose was di †erent between the silages. Mammary extraction rates of glucose, BHB, histidine, leucine and lysine were a †ected by their arterial concentrations. The interaction between silage fermentation and casein infusion was signiÐcant (P \ 0É05) in arterial concentration of branched-chain amino acids, the increase being greater with FA-silage than with LAB-silage. The results suggest that the supply of glucose may limit milk synthesis in cows given restrictively fermented grass silage and barley, whereas the supply of amino acids may be more limiting with extensively fermented silages. It is further suggested that the mammary gland has a capacity to adapt itÏs efficiency of nutrient uptake according to the supply and mammary demand of nutrients.