Aceh's MSME population is engaged in various types and different economic activities with different incomes. The research that has been done in analyzing the income of MSME owners has been widely carried out, using simple regression analysis, the results show that capital and labor have a significant effect on the amount of income. Through the classification of MSME data sets, simple regression analysis can be improved by considering the relationship with different MSME classifications and regions. Aceh MSME data has a hierarchical data structure with individual levels nested in the MSME classification group nested in regional groups (districts/cities). So, is a multilevel regression analysis. This study was conducted to determine the effect of the independent variables at level 1 (type of business, labor and capital), level 2 variance (classification of MSMEs), and level 3 variance (district/city) on the income of Aceh MSMEs. The Aceh MSME data has very varied data forms and contains outliers, so a robust standard error is needed that is not sensitive to violations of the assumptions in this study. The model formed in the Aceh MSME data is a 2-level model with the addition of independent variables. The results obtained indicate that the MSME classification variance and the variable x_1 (type of business, namely trade, livestock and transportation), variable x_2 (labor), and variable x_3 (capital) affect the income of MSMEs in Aceh Province. Diversity at level 1 (individual MSMEs) only has an effect on MSME income by 29%, while after being given diversity at level 2 (MSME classification) it will have an effect on MSME income by 71%.