Abstract: African Horsesickness (AHS) is a controlled disease in South Africa. The country is divided into an infected area and a control area. An outbreak of AHS in the control area can result in a ban of exports for at least 2 years. A retrospective epidemiological study was carried out on data collected during the 2004 AHS outbreak in the surveillance zone of the AHS control area in the Western Cape Province. The objective of this study was to describe the 2004 outbreak and compare it with the 1999 AHS outbreak in the same area. As part of the investigation, a questionnaire survey was conducted in the 30 km radius surrounding the index case. Spatial, temporal and population patterns for the outbreak are described. The investigation found that the outbreak occurred before any significant rainfall and that the main AHS vector (Culicoides imicola) was present in abundance during the outbreak. Furthermore, 63% of cases occurred at temperatures < or = 15 degrees C, the Eerste River Valley was a high risk area, only 17% of owners used vector protection as a control measure and 70% of horses in the outbreak area were protected by means of vaccination at the start of the outbreak. The study revealed that the current AHS control measures do not function optimally because of the high percentage of vaccinated horses in the surveillance zone, which results in insufficient sentinel animals and the consequent failure of the early warning system. Alternative options for control that allow continued export are discussed in the paper.