Abstract The incidence rate of clinical mastitis in 125 herds with a low annual bulk milk SCC (<150,000 cells/ml) was modeled using a Poisson regression model. The rate of clinical mastitis was significantly associated with some variables that increased the exposure to environmental microorganisms: poor cubicle cleanliness increased the mastitis rate; rubber mats in cubicles were associated with a higher incidence; and drinking water from sources other than public water also increased the rate of mastitis. Other variables may be associated with host resistance: an increasing percentage of cows leaking milk increased the rate of mastitis; postmilking teat disinfection was associated with a higher incidence of clinical mastitis; and a high frequency of cubicle disinfection was also associated with more mastitis. Three other variables were associated with the rate of mastitis: breed (Holstein-Friesian had a lower incidence than the Meuse-Rhine-Yssel breed); use of sugar beet pulp in the ration increased the mastitis rate; and in herds with high milk production a higher incidence of mastitis was observed. These items are discussed in respect to their causal relation to clinical mastitis.