This comment raises three main issues about He and Qin’s (2004) attempt at modeling investment in the PRC. The first is this author’s skepticism about the general applicability of the neoclassical model of investment to the PRC. Second, that their model for business investment, based on the neoclassical theory of investment, can be viewed as an approximation to an accounting identity derived by manipulating two other identities, namely, that of the capital share in output, and that of the motion of the capital stock. It is shown that the difference between He and Qin’s equation and the identity is simply that they use the rental price of capital, while the identity relies on the profit rate. At best, all their analysis would indicate is that rental price of capital and profit rate are different. It is also argued that the empirical results are not clearly related to the supposed theoretical model. Based on this, the conclusion is that the policy implications of He and Qin’s alleged model are somewhat dubious. Third, He and Qin’s equation for government investment introduces the deviations of output from the long-run trend as an explanatory variable, estimated using an aggregate production function. The problems underlying this latter concept make the estimation of the output trend using this method a questionable exercise. Also, the empirical results suffer from serious problems of interpretation.