Raw milk was analysed before and after microfiltration with the aim to reduce its somatic cell count, carried out with commercially available microfiltration equipment, in order to investigate further possible changes of the milk composition. The objective was to find indicators that could serve as proof of this illegal on-farm procedure in case of milk delivery to dairies. The effect of microfiltration was clearly visible, with a somatic cell reduction rate of approximately 75%. Out of several milk constituents tested, only the free fatty acids content showed a highly significant increase in microfiltered raw milk. Furthermore, in raw milk with somatic cell counts > 400 000/ml, the corresponding N-acetyl-beta-D-glucosaminidase (NAGase) activity remained high, even after the cell count was reduced by microfiltration. For a suspicion that somatic cell count in raw milk has been lowered by microfiltration, a combination of criteria appears to be a possible strategy: daily milk delivery > 200 1, somatic cell counts < 400 000/ml, gliding median values of free fatty acids in six consecutive routine analyses of raw milk over a period of two months > 0.4 mmol/l, and NAGase activity values of > 1.2 mu mol 4-methylumbelliferone/ml min.