A paper in this JOURNAL by Blitzer, Dasgupta and Stiglitz (I98I), hereinafter BDS, entitled 'Project Appraisal and Foreign Exchange Constraints', examines the implications of various non-optimal macroeconomic adjustment policies for project evaluation.' The discussion is presented in terms of a simple two good, one consumer economy with price distortions which lead the ratio of the domestic price of the exported good (E) to the domestic price of the imported good (I) to be greater than the internati-onal price ratio. It is argued that the resulting trade gap will force the government to pursue an adjustment policy which may take the form of taxation of income or commodities, rationing or foreign borrowing; and it is shown that the choice of policy will determine the appropriate shadow price formulae to be used in project appraisal. The view that world prices provide shadow prices for traded goods is argued to be correct for some but not all of the macroeconomic policies considered. This comment argues that the BDS analysis fails to specify how tax revenue is to be incorporated in the model. As a result, the various policy options open to the government are not clearly brought out. We set out below, in Section I, an amplified version of the BDS general equilibrium model which includes tax revenue, and use it to show how the different adjustment mechanisms have to be related to the public sector budget constraint. In Section II we reconsider their analysis and show that if either an income or a commodity tax on the importable is used to ensure public sector balance, then world prices remain shadow prices for traded goods. In Section III we make the point that the presence of import tariffs does not imply, in the context of a real general equilibrium model, that the balance of payments constraint will be violated. The conclusion is drawn that 'endogenous foreign borrowing' cannot be analysed without respecifying the basic BDS model. Our comment is confined to the problem of shadow pricing traded goods in the BDS model; we do not consider rationing as a means of macroeconomic adjustment policy, nor do we comment on their Section III where they analyse the additional problems which follow from introducing a non-traded commodity into the model.