In the literature on the application of STIRPAT to environmental impacts of population and affluence, the parameter estimates differ from study to study. One example is the effect of population size on CO2 emissions, which is concluded to be very close to 1 in some studies (e.g., York et al., 2003) while far from 1 in others (e.g., Shi, 2003). What can explain these differences in results? In the present paper, I offer an alternative model equivalent to STIRPAT, which explicitly specifies the different role of technology (T) in STIRPAT from IPAT. By the alternative model, I conclude that different functional forms of STIRPAT can be one explanation for the difference among estimates in the studies on environmental impacts of population and affluence. The alternative model can also help to determine which factors to be added in STIRPAT.