In the present study a murine encephalitis model was used to investigate if protection against neosporosis could be achieved by immunisation. Groups of 10 mice were immunised with a sublethal dose of live Neospora caninum tachyzoites, N. caninum antigens incorporated into iscoms, N. caninum lysate mixed with Quil A, or N. caninum lysate in PBS. Control mice were given Quil A only. Challenge infection with 2.5x10(6) N. caninum tachyzoites resulted in clinical symptoms that remained until the end of the experiment in the controls. In contrast, mice immunised with live parasites or parasite lysate in Quil A only showed mild and transient symptoms. Of nine mice immunised with N. caninum iscoms, seven recovered while two died. Most severely affected were the mice immunised with parasite lysate only; all of them died within 28 days post-infection. Histological examination and scoring of brain lesions gave a significantly lower (P<0.0001) lesion score in mice immunised with live parasites than in controls. The groups immunised with iscoms or lysate and Quil A also had reduced lesion scores (P<0.04 and 0.07, respectively) but not the group given parasite lysate alone. The lesions seen in the latter group differed from those in the other groups. There was less cellular reaction and more tachyzoites indicating an active infection. The N. caninum specific antibody responses and cytokine production (IFN-gamma, IL-4 and IL-5) of splenocytes were analysed at the time of challenge infection. The results suggest a correlation between protection and high levels of IFN-gamma. Also, the immune responses recorded in mice immunised with parasite lysate without adjuvant were relatively weak and more towards the Th2 type, when compared with the other immunisation schedules. This is consistent with the weaker inflammatory response observed in the brains of these mice.