Blood samples were collected from 3449 cows on 57 representative Ontario dairy herds during the summer of 1998 and analysed for antibody to Neospora caninum using an ELISA. Forty-eight herds (2742 cattle) contained at least one N. caninum-seropositive animal. Two composite milk samples were collected from all cattle: the first on the day of blood collection and the second 68 to 365 days later. All milk samples were submitted for bacteriological culture. Ontario Dairy Herd Improvement Corporation (DHI) data were available for 3162 cattle in the 57 herds at the time of bleeding. Furthermore, complete DHI data were available for 1658 cattle that were culled between 12 and 24 months following blood collection. Using a standardised ELISA sample-to-positive (S/P) cut-off of ≥0.45, the corrected seroprevalence was 8.2% overall and 10.1% within seropositive herds. At blood collection the odds of N. caninum-seropositive cows having a high linear score (≥4.0; equivalent to a somatic cell count ≥200,000 cells/ml) was 27% less than for seronegative animals. Similarly, at the time of culling, the odds of having a high linear score was 22% less in N. caninum-seropositive cattle. Overall, linear score was lower in N. caninum-seropositive cattle at culling. After controlling for herd, parity, days in milk, and the interval between collection of milk samples, the odds of N. caninum-seropositive cattle testing positive for an environmental pathogen (i.e. environmental Streptococcus species and coliforms) on the second milk sample was 56% less than for seronegative animals. The odds were 83% less at a higher ELISA S/P cut-off of ≥0.70. Finally, the odds of N. caninum-seropositive cattle developing a new infection with a major pathogen (environmental or contagious) were 60% less than seronegative cows using the higher ELISA S/P cut-off.