The possible direct relationship between climate variations and abortion in Neospora caninum-infected cows has not been studied. The objective of this study was to determine whether climate changes could be a risk factor for abortion in N. caninum-infected cows, and was based on yearly serological screening for neosporosis and on the confirmation of N. caninum infection on aborted fetuses in two high-producing dairy herds with a mean 27% seroprevalence of N. caninum antibodies. The final population study was comprised of 357 pregnancies in seropositive animals. Logistic regression analysis indicated no significant effects of herd, N. caninum antibody titre, climate variables during the first and third trimesters of gestation, mean and maximum temperature– humidity index values during the second trimester of gestation, and previous abortion on the abortion rate. Based on the odds ratio, a 1-unit increase in lactation number yielded a 0.85-fold decrease in the abortion rate. The likelihood of abortion was 1.9 times (1/0.54) lower for pregnant cows inseminated with beef bull semen compared with Holstein-Friesian bull semen. The likelihood of abortion decreased significant and progressively by factors of 0.5, 0.41 and 0.3 for the respective classes 40–49, 30–39 and <30 rainfall mm during the second trimester of gestation (using the class ‡60 rainfall mm as reference). As a general conclusion, it seems that increased rainfall in a dry environment can compromise the success of gestation in N. caninum-infected cows. Attempts should therefore be made to reduce environment effects during the second trimester of gestation, a period in which the immune response of cows is diminished.