This paper empirically investigates whether there are asymmetries in the responses of US government tax revenue and expenditure to debt levels and economic conditions over the business cycle. State of the art regime switching regression models, including Threshold Regression and Markov Switching, are investigated. Both sides of the government budget show asymmetries, but the asymmetries for tax revenue show greater statistical significance. The results show that both tax revenue and expenditure respond to high debt levels, with the asymmetry in this response showing that fiscal authorities take weaker action in response to debt during poor economic times. In addition, the asymmetric response to economic conditions for both sides of the budget shows that stronger countercyclical policy is taken during poor economic times.Keywords: asymmetric fiscal policy, fiscal policy sustainability, cyclicality, government expenditure, tax revenue Recently a distinct but related body of work has investigated whether the effect of fiscal policy shocks vary over the business cycle. For example, Auerbach and Gorodnichenko (2012) focus on the size of spending multipliers in 1 It is useful to note that both implementation and legislative lags for fiscal policy are considerably longer than for moneraty policy. Hence, we expect a larger degree of persistance in government spending and tax revenues.