Theory: The political business-cycle theory is used to test for state-level accelerations in contract awards and deobligations induced by presidential elections. Hypotheses: Government contracting agencies accelerate contract awards, and reduce the level of deobligations, immediately before presidential elections. The cycles should not involve any total increase in spending. Methods: Pooled-time-series analysis of monthly Department of Defense and civilian agency contract awards and deobligations in each state between October 1985 and December 1992. Results: Both contract awards and deobligations behave as expected. Civilian agency contract awards rise immediately before presidential elections and primaries, and both Department of Defense and civilian agency deobligation activity virtually cease as the elections approach. Although the total level of activity is unchanged in election years, the cycles are clear. This supports the notion that the cycles are important politically even though they have little macroeconomic impact.