The article examines how to accurately measure volatility in African stock markets by considering changes in volatility levels over time. By incorporating these shifts into standard models, the study shows that volatility persistence and long memory are often overestimated without accounting for these changes. The research suggests that allowing for time variation in volatility improves forecasting performance for some African markets. This study is one of the first to incorporate endogenously determined regime shifts into volatility estimates, shedding light on the impact of structural breaks on volatility in African stock markets.