Informational efficiency is a fundamental pillar of well-functioning financial markets, as it underlies informed investment decisions, effective risk management, and broader economic stability, particularly in emerging African markets, where inefficiencies are more likely to persist. This study assesses the weak-form informational efficiency of six major African stock markets – Johannesburg, Casablanca, Botswana, Nigeria, Egypt, and the Regional Stock Exchange – through the lens of long-memory behavior in returns and volatility. This is achieved by employing four advanced models: ARFIMA-FIGARCH, ARFIMA-FIEGARCH, ARFIMA-FIAPARCH, and ARFIMA-HYGARCH. Each of these models is specifically designed to capture long memory in both the conditional mean and variance. The empirical results demonstrate that the ARFIMA-FIGARCH framework, across all four model variants, consistently outperformed alternative specifications in fitting the return and volatility dynamics of all six African stock market indices. The estimated fractional differencing parameters in both the mean (dARFIMA) and variance (dFIGARCH) equations were highly statistically significant at the 1% level for each market, confirming the presence of persistent long-memory behavior. This strong evidence of long-range dependence implies that past return information is not fully reflected in current prices, thereby violating the assumptions of weak-form market efficiency. Consequently, these findings provide compelling and systematic evidence against the weak-form Efficient Market Hypothesis (EMH) for the markets studied, highlighting a common structural inefficiency across the African financial landscape.