Attempts were made to transmit Chlamydia psittaci, the causal agent of enzootic abortion of ewes, in three different ways. Ten ewes were inseminated artificially with freshly collected semen containing 10(5) CELD50 chlamydia. Serological evidence of infection was found two weeks before parturition in nine ewes and the organism was recovered from three of them. By six weeks post partum antibody titres had fallen and were negative in six ewes. Twenty-four hours after service two groups of 10 ewes were infected intravaginally with 10(8) CELD50 and 10(3) CELD50 chlamydia respectively. Positive complement fixation titres were present in the first half of pregnancy in all the ewes in the high dose group but not the low dose group. None of the ewes showed evidence of infection at parturition. Fourteen ewes were served by four rams which had been intravenously infected with 10(8) CELD50 chlamydia four to six days earlier. Following service seroconversion occurred but titres became negative again by late pregnancy. No microbiological evidence of infection was detected in any of the ewes at parturition but complement fixation titres were positive in 12 of 14 ewes sampled six weeks post partum. The 14 ewes were sampled during their pregnancies the next year and none showed any evidence of chlamydial infection. It is concluded that venereal transmission of C psittaci is biologically feasible in sheep, but that under normal systems of flock management in Britain it is unlikely to contribute greatly to the epidemiology of enzootic abortion of ewes.