Rosellinia spp. cause substantial mortality of cocoa (Theobroma cacao), especially in acidic soils rich in organic matter (O.M.) and, reportedly, low available phosphate. Cultural and chemical control are relatively ineffective and a supplementary biocontrol strategy would be highly desirable. Our aim was the isolation and evaluation of promising mycoparasites against Rosellinia of cocoa. We isolated 45 mycoparasites using a baiting technique. Three of them and a commercial isolate were prioritized for further evaluation based on their broad hostrange and/or effectiveness in pot bioassays. Because Rosellinia is favored by the same soil factors as mycoparasites, we investigated the effect of soil pH (range 4-6), O.M. (6-12%), and phosphate (0-1000mgkg_1) on their interaction. In vitro, both pathogens and antagonists grew well on a wide range of phosphate (0-9500mgkg_1) and at various pH (3.0-10.0) levels. The pH optimum usually coincided with the soil of origin and the phosphate optimum depended on the phosphate source used. In pot bioassays using natural soils, mixtures of antagonists exhibited more control than single strains and mixtures were effective over a wider range of soil conditions. High O.M. favored the pathogen more than its biocontrol agents. Soil pH interacted with the treatments such that biocontrol was most efficient at higher pH values and with the highest mixtures of antagonists. The addition of phosphate enhanced plant growth but had little effect on biocontrol efficacy. We conclude that mixtures of mycoparasites should be field tested in combination with liming and removal of O.M.