Few empirical studies have estimated the direct effects of biotic (disease and pest) and abiotic (e.g. drought and flood) stresses on cocoa-producing households. As such, this study extends the existing literature by using household-level data from Ghana over three cocoa growing seasons (2002, 2004 and 2006) in a regression framework to estimate the responsiveness of cocoa yields to biotic and abiotic stress at the household level. The results show that, for farms exposed per year, overall stress from pests makes up the highest percentage, followed by disease stresses and abiotic stresses. In addition, the results from the regression model show that cocoa yields decline by 0.046%, 0.013% and 0.003% respectively for every one percent increase in the proportion of the farm affected by disease, pests and abiotic stress that persist for a year. The findings of this study suggest that the government of Ghana should consider expanding the scope of the National Cocoa Diseases and Pest Control Programme to include other pests that are not included in the programme. We also recommend an insurance product for cocoa to help farmers manage the risks of abiotic stresses such as droughts and floods that destroy investments and potential income.