Finite-state models of control systems were proposed by several researchers as a convenient mechanism to synthesize controllers enforcing complex specifications. Existing techniques for the construction of such symbolic models have so far relied on certain stability or stabilizability assumptions. In this paper, we show that these assumptions can be relaxed and prove that large classes of unstable systems admit symbolic models. The effectiveness of the proposed results is illustrated by synthesizing a controller for an inverted pendulum subject to a schedulability constraint.