An interlaboratory comparison on cheese making trials was conducted to examine the differences between model cheeses manufactured in pilot plants at six European laboratories. The experimental design (within each laboratory) for the model cheese was an unreplicated 2(3) full factorial design in one block of eight cheese vats. The three factors were the pasteurization of skimmed milk, microfiltration of skimmed milk and pasteurization of cream. Although the manufacture, sampling and analyses were generally standardized the gross composition and sensorial properties of the cheeses differed greatly across the laboratories. Good statistical design and analysis of the experimental data aided in minimizing the effect caused by the technical difficulties experienced by some of the laboratories during the cheese making trials Despite the problems encountered, all sir laboratories noted similar effects between model cheeses that were produced from milk in which the initial load of indigenous flora was reduced by either pasteurization or microfiltration. The cheeses produced from milk in which the raw milk flora had been reduced had lower concentrations of D-lactate and a less intense aroma and odour. With few exceptions the individual laboratories observed similar effects, although these were less significant when compared with the overall effect. (C) 1998 Academic Press