Worldwide, medical errors occur at an unacceptable rate. Fortunately, various steps can be taken to minimise medical errors, including submitting reports to incident reporting systems, following which incidents can be categorised using error classification systems to inform risk management strategies. To this end, a diverse range of classification systems have been developed for different healthcare settings, but primary care in Malaysia has received limited attention. This study was carried out to develop an error classification system for categorising the type of medical errors in Malaysian primary care by integrating and modifying two primary care error classification systems. The new error classification system comprised of eleven major categories that branched into a total of 60 lower categories. After 107 participants analysed three cases studies, the inter-rater reliability of the 11 major categories were found to be low (Krippendorff‘s alpha = 0.347). Unsurprisingly, the 60 lower categories had an even lower inter-rater reliability (Krippendorff‘s alpha = 0.143). The participants then provided feedback about the error classification system by completing a modified questionnaire. Most of the participants viewed the error classification system as easy to use and understand. Furthermore, most of them would be willing to use the error classification system again in the future. In summary, this study successfully developed an error classification system for Malaysian primary care. Nevertheless, its low level of reliability would need to be addressed before the system can be adopted by primary care providers.