This paper studies the long-term effects of government procurement on small business growth, using a novel matched dataset that links federal contracts awarded to small businesses and bidding data from construction procurements with firm-level outcomes from restricted Census administrative data. Using a matched difference-in-differences approach and a winner-versus-losers design, the study documents several findings. First, procurement contracts lead to sustained increases in small business revenue, employment, and earnings, with effects persisting for up to five years beyond the contract period. These effects are not primarily explained by repeat contracting or long-duration contracts. Firms also experience short-run crowding out of private sales, which later recover and exceed pre-contract levels. Using micro-level credit scores and labor share data, the study shows that financially constrained and capital-intensive firms exhibit stronger post-contract growth. A dynamic investment model with financial frictions and government contracting demonstrates that a one-time procurement shock can rationalize these effects through a combination of productivity gains and relaxed financial constraints, thereby enabling irreversible investment. Counterfactual simulations indicate that this persistence primarily reflects productivity gains rather than short-run easing of financing constraints.