The effects of increased somatic cell count, whether caused by infection or by experimental infusion of bacterial endotoxin, on the distribution in milk of caseins between the micellar and soluble forms were investigated. The relationship of somatic cell count to some cheese-making parameters was also studied. With quite modestly elevated cell counts (2–3 × 106/ml) increases of up to 37% in total casein in the soluble phase were observed, most of which was contributed by β-casein, while κ- and αs1-caseins increased only slightly. With storage at 4°C, the concentrations of all the caseins, Ca and phosphate in the soluble phase increased substantially during the first 48 h, but this was followed by a slight decline on further storage. Rennet clotting time, losses of fat in whey, curd moisture, and losses in curd yield and rigidity were all greater the higher the somatic count. Clear differences were detectable in these parameters between milks of very low cell count (e.g. 5 × 104 cells/ml) and milks with counts more typical of those found in bulk supplies (e.g. about 5 × 105 cells/ml). If these findings can be reproduced in commercial practice even a modest reduction in bulk milk somatic cell counts might be expected to bring definite benefits.