Studying spatial and temporal plant disease dynamics helps to understand pathogen dispersal processes and improve disease control recommendations. In this study, three cacao plots devoid of primary inoculum of Phytophthora megakarya (causal agent of cacao black pod rot disease) upon establishment in 2006 were monitored for presence of disease on a weekly basis from 2009 to 2016. Ripley's K(r) function, join count statistics and Fisher Exact test were used to analyse spatial and temporal disease dynamics. Disease distribution maps showed aggregated disease patterns in all plots although for the years of disease onset, exogenous primary infections were mostly randomly distributed. The K(r) function confirmed these results indicating that inoculum generally disperses only over short distances. Moreover, significant positive spatial autocorrelations showed that diseased trees were often clustered up to a distance of 3-9 m. Temporal disease progression was low, meaning that endogenous inoculum failed to establish itself which is partly explained by rigorous phytosanitation and partly by unfavourable microclimatic conditions for disease development. Since P. megakarya had difficulty establishing itself in the plots, proximity to already infected cacao plantations drove infection dynamics. Thus, isolation of newly established cacao plantations from infected ones and rigorous phytosanitation as a preventive strategy appears to be an effective approach to control black pod for newly established cacao plantations.