The effects of dexamethasone and promethazine on the amelioration of pulmonary oedema in East Coast fever were investigated. The clinical effects of these drugs were further investigated when used in conjunction with the antitheilerial drug, buparvaquone. In the first experiment, 15 crossbred (Friesian x Zebu) steers were divided into four groups. With the exception of the animals in group IV, that served as a control group all the others were infected with Theileria parva sporozoites. On the second day of the febrile reaction, the steers in groups I and II were treated with dexamethasone (0.1 mg/kg) and promethazine (1 mg/kg), respectively. Group III steers served as the infected untreated controls. On the fifth day of the febrile reaction the animals in groups I, II and III were infused intravenously with tattoo ink suspension and 1 h later sacrificed for post-mortem examination and tissue sampling. The clinical picture indicated that both drugs significantly mitigated dyspnoea and the post mortem examination revealed a significant reduction in morphological changes. Tattoo ink particle count reflected a significant (P< 0.01) reduction in vascular leakage in the treated animals, with promethazine being significantly (P < 0.05) more effective than dexamethasone in this respect. In the second experiment, 18 steers were infected with T. parva sporozoites, and then were randomly allotted into three groups each of which contained six animals. After the onset of ECF clinical signs, the animals in the first two groups were treated with buparvaquone in combination with either dexamethasone (group I) or promethazine (group II), and the third group was treated with buparvaquone alone. The results indicated that all the animals in groups I, II and III recovered well and no significant differences were observed in clinical disposition between the groups. Two months later, serum samples were collected from the refractory animals and demonstrated the presence of antibodies against T. parva. When the animals were subsequently artificially challenged with T. parva, none of them succumbed to clinical disease. The same T. parva stabilate stock was used in both experiments and it proved to be infective in a separate batch of steers.