The successive steps of an integrated analytical procedure aimed at the accurate determination of butterfat fatty acid composition, including trans-18:1 acid content and profile, have been carefully checked. This sequential procedure includes: dispersion of a portion of butter in hexane/isopropanol (2:1, vol/vol) with anhydrous Na2SO4, filtration of aliquots of the suspension through a microfiltration unit, subsequent preparation of fatty acid isopropyl esters (FAIPE) with H2SO 4 as a catalyst, and analysis of total FAIPE by capillary gas-liquid chromatography (GLC). Isolation of trans-18:1 isomers was by silver-ion thin-layer chromatography (Ag-TLC), followed by extraction from the gel of combined saturated and trans-monoenoic acids with a biphasic solvent system. Analysis of these fractions by GLC allows the accurate quantitation of trans-18:1 acids with saturated acids (14:0, 16:0, and 18:0) as internal standards. A partial insight in the distribution of trans-18:1 isomers can be obtained by GLC on a CP Sil 88 capillary column (Chrompack, Middelburg, The Netherlands). All steps of the procedure are quite reproducible, part of the coefficients of variation (generally less than 3%, mainly limited to butyric and stearic acids) being associated with GLC analysis (injection, integration of peaks) and, to a lesser extent, to FAIPE preparation. FA1PE appear to be of greater practical interest than any other fatty acid esters, including fatty acid methyl esters (FAME), for the quantitation of short-chain fatty acids, because peak area percentages, calculated by the integrator coupled to the flame-ionization detector, are almost equal (theoretically and experimentally) to fatty acid weight percentages and do not require correction factors. With this set of procedures, we have followed in detail the seasonal variations of fatty acids in butterfat, with sixty commercial samples of French butter collected at five different periods of the year. Important variations occur around mid-April, when cows are shifted from forage and concentrates during winters spent in their stalls to fresh grass in pastures. At this period, there is a decrease of 4:0-14:0 acids and of 16:0 (-2 and -6%, respectively), while 18:0 and cis-plus trans-18:1 acids rise suddenly (2 and 5%, respectively). These modifications then progressively disappear until late fall or early winter. Other variations are of minor quantitative importance. Although influenced by the season, the

fat -extraction -isopropyl esters -gas Iiquid chromatography -response factor -low caloriespread

Separation and quantitation of triacylglycerols tricaprylin to tristearin was achieved by high‐performance liquid chromatography with refractive index detection and an eluent composed of propionitrile and butyronitrile (80:20, vol/vol). Peak identification was based on the logarithms of retention times, and quantitation was achieved by way of theoretical relative response factors. The validity of the response factor calculations was tested by analysis of primary standard mixtures of saturated triacylglycerols and of triacylglycerols obtained by interesterification of binary mixtures of monoacid triacylglycerols.