This paper examines whether the introduction of government consumptionexpenditure in a standard one good model of the international real businesscycle is sufficient to reconcile the theory with the existing pattern ofinternational consumption and output correlations. I calibrate the model totwo different pairs of countries and generate the simulated distribution ofconsumption and output correlations implied by several specifications of themodel. It is shown that the model can account for existing internationalconsumption correlations only under very specific assumptions about the sizeof effect of government expenditure on agents' utility or the variabilityof government expenditure shocks. Crucial parameters are identified and thesensitivity of the results discussed.