We investigate the problem of exact boundary controllability of semilinear one-dimensional wave equations. We prove that it is possible to move from any steady-state to any other one by means of a boundary control, provided that they are in the same connected component of the set of steady-states. Full paper.