An epidemiological study was conducted on 20 dairy herds previously diagnosed as seropositive for Neospora caninum. The number of females per farm varies from 41 to 296. All females present on the farms were bled once in the period of July and August 2000. A total of 3002 females were bled. An indirect ELISA was used to determine the serostatus of the animals. The analysis of the data was performed in four steps: (1) descriptive statistics about the serological status and general characteristics of the cattle; (2) calculation of vertical and horizontal transmission; (3) an univariate analysis and, (4) a multivariate logistic regression analysis with herd as random effect. The within-herd seroprevalence varied between 25.0% (34/136) and 70.5% (203/288). Seven hundred and forty-seven dam-daughter pairs were available, involving daughters of any age. Daughters in the specific age-class of 2- and 3-years old had a higher seroprevalence (P<0.01) compared with younger and older age-classes. The risk of being seropositive when being born to a mother that tested seropositive (prevalence ratio (PR)) was 2.8-fold increased which coincides with a 5.3-fold increased odds. The probability of horizontal infection amounts to 0.22. The probability of a seropositive offspring due to vertical transmission was 0.64 (attributable fraction among exposed (AFexp)). The multivariate logistic regression showed a significant 6.0-fold increased odds for being seropositive when born from a seropositive mother. Also the within-herd seroprevalence level was significantly associated with the serostatus of the daughters. In the specific conditions of the dairy herds involved in this study, the serostatus of the cows should be not used as a predictor of the serostatus of daughters due to the increased probability of horizontal transmission.