The increase in payment of afforestation incentive program recently indicates a high opportunity cost of the forestry works in Taiwan. The purpose of this study is to use forestry censuses data to examine empirically the factors to affect income of the forest households in central region of Taiwan. An analysis model is developed mainly based on property right theory to find out the factors. Application of spatial statistical methods to detect the spatial effect of different townships, and use regression and quantile regression methods to analyze what factors would affect forest income. Based on regression analyses, the study result shows forest household income is affect positively by the size of forestland, the number of labours and private forest ownership of land, but the proportion of self-employed labours negatively. The marginal effect in different forestry income would not the same.