In this paper, we examine the statistical properties of rainfall data and temperature in six sub-Saharan African countries in the western, eastern, and southern regions (Botswana, Ethiopia, Ghana, Nigeria, Uganda, and South Africa) using time series data spanning between 1900 and 2012. By using linear trends, seasonality, and long-range dependence models, in fractional or I(d) frameworks, the results first indicate that time trends are required in most cases to explain the time series properties of the climatic series. Evidence of anti-persistence (d<0) or I(0) behavior is found for the rainfall data, while long memory (d>0) is found for the temperature data. Evidence of structural breaks are only found in the cases of Ethiopia, Ghana, and Uganda for the temperature data. With both series displaying significant evidence of seasonality and by working with the seasonally differenced data, the results show evidence of I(0) behavior or anti-persistence (d<0) for the rainfall data but long memory (d>0) for the temperature data. Testing the causality between the two variables, the results indicate evidence of causality in the two directions in all cases except for the case of the temperature on the rainfall in South Africa. The implication of the results obtained here is that erratic or constant rainfall is expected in Africa in the future while temperature is likely to continue to increase, and these subsequently lead to future warming experiences.