Abstract Morbidity and mortality associated with an infectious disease outbreak can be mitigated by early detection followed by swift action. Modelling, tracking and predicting disease outbreaks are therefore priorities for public health agencies. New media data sources, including social media platforms, the internet and mobile phone applications, now aid in detecting outbreaks earlier than would have been possible using traditional surveillance methods alone. I review the literature on uses of new media methods for detecting disease outbreaks in humans and animals, with a focus on veterinary diseases and the difference in challenges compared with human disease surveillance. I then discuss the complex issue of evaluation of new media-based surveillance systems. The proliferation of new media methods for disease surveillance has not included published evaluation of each method or of the challenges faced, which limits the potential for a particular method to be applied outside its original context.