Whether government spending can boost the pace of economic growth is widely debated. In the neoclassical growth model, it is supplies of productive resources and productivity that determine growth in the long-run. In endogenous growth models, an increase in government spending may raise the steady-state rate of growth due to positive spillover effects on investment in physical and/or human capital. This paper examines the relationship between government spending and non-oil GDP in the case of Saudi Arabia. Using time-series methods and data for 1969-2005, we find that increases in government spending have a positive and significant long-run effect on the rate of growth. Estimated effects of current expenditure on growth turn out to exceed those of capital expenditure -- suggesting that government investment in infrastructure and productive capacity has been less growth-enhancing in Saudi Arabia than programs to improve administration and operation of government entities and support purchasing power. We discuss possible reasons for this finding in the Saudi case and draw some policy implications.