The first large-scale election observation was of Zimbabwe’s 1980 independence elections. Since then, election observation has become a regular worldwide feature and many international organisations, official agencies, and non-governmental organisations field observation teams. They all use similar methodologies, largely derived from the original 1980 model. A third of a century later, it may be time to consider whether the use of electoral observation has outlived its usefulness – or is itself being used to mask forms of electoral cheating. This paper considers five 21st century African elections – in Kenya (2007), Zimbabwe (2008, 2013 and 2018) and Zambia (2016), through the reflections of a pioneer of the 1980s observation prototype