Repeated trunk injections of tetracycline hydrochloride and oxytetra‐cycline over a 27 month period, retarded the rate at which clove trees in Zanzibar, died from sudden death (SD) disease. This result provides additional support for an MLO aetiology for SD, but indicates that the tetracycline formulations and method of application, as described here, is not a feasible method of SD control in the field. Bioassay of leaf discs and root sections using agar plates seeded with Bacillus subtilis, at various times after injection, indicated that antibiotic distribution to the canopy was extensive, but that down to the roots was minimal. The proportion of leaf discs showing tetracycline activity declined over a 54 day period post‐injection, from 75,% to 17%. Mean tetracycline levels in leaves per tree peaked at 2.4 μg/ml, 21 days after injection. The highest individual leaf disc level was 7 μg/ml. Most trees maintained concentrations of tetracycline inhibitory to B. subtilis growth for 2,1–41 days.