The aim of this study is to assess the validity of EKC and PHH empirically in relation to carbon dioxide, methane and nitrous oxide emissions in relation to per capita income, using spatial econometric regression on data from 40 African countries (1971–2019). The paper also examines the impact of foreign direct investment, electricity consumption and trade openness on the aforementioned atmospheric pollutants. The paper supports the EKC theory: as economies grow, environmental degradation can decrease once a certain threshold is reached. According to the SAR and SEM models, trade openness has a negative impact on environmental quality and is strongly correlated with emissions of carbon dioxide and nitrous oxide. Because African economic unions engage in a lot of trade, this is consistent with non-spatial outcomes. The negative correlation between these pollutants and electricity usage points to a move toward clean energy. As a measure of economic growth, per capita income has a positive correlation with every pollutant. In contrast, the income square for methane and nitrous oxide in SAR and SEM models is negative, suggesting that environmental conditions initially improve. Furthermore, in line with earlier models, the SARAR model indicates that every contaminant is significant. Interestingly, in contrast to other models, methane coefficients are considerable. The environmental Kuznets curve for methane and nitrous oxide is supported by the income squared variable, which is negative and significant for all pollutants. Except for FDI, which is the sole non-significant variable in the Durbin model, all of the variables are significant. This does not align with the proposals of the pollution hypothesis. In this sense, the study shows that the validity of EKC and PHH depends on factors such as modeling approach, variables, spatial dependencies and study duration.