The protection afforded by an experimental, killed, adjuvanted vaccine derived from the A22 strain of Chlamydia psittaci (ovis) against ovine enzootic abortion was studied. The vaccine was used undiluted (group A), at a dilution of 10(-3) (group B) and at a dilution of 10(-6) (group C). A fourth control group (group D) was inoculated with all components of the vaccine except the chlamydial antigen. A group of rams (group R) was also vaccinated with the chlamydial antigen diluted to 10(-3). Animals were challenged 70 days after mating with the A22 strain of C. psittaci (ovis) and were studied throughout pregnancy and the subsequent lambing period. Their cell-mediated immune responses were examined using a skin test and their humoral immune responses were studied using an ELISA. Tests for excretion of chlamydiae in their faeces and genital tract during pregnancy and after parturition and in the faeces of their lambs were made. The reproductive performance of the ewes was assessed by calculating the average weight of lambs produced per ewe in each group. The experimental vaccine protected the ewes in groups A and B against challenge with C. psittaci (ovis) as none showed clinical signs of OEA or excreted chlamydiae. The average weight of lambs produced per ewe in both groups was greater than 4 kg. Both groups seroconverted after vaccination but not all of them were positive to the skin test. The experimental vaccine at 10(-6) dilution of antigen did not protect the ewes as three of 10 ewes displayed clinical OEA and excreted chlamydiae in the products of parturition.(ABSTRACT TRUNCATED AT 250 WORDS)