The near-field motion on the surface of a uniform half-space caused by the passage of the rupture on a vertical, strike-slip fault has been studied by the use of a dislocation model. The fault is modeled by an infinitely long-buried dislocation of finite width; rupture propagates horizontally along this fault and past the observation points with a constant rupture velocity lower than the Rayleigh wave velocity. Peak amplitudes caused by the passing rupture front are primarily controlled by the depth of the top of the fault and the rupture velocity, when the slip on the fault and the rise time are held constant. A nonvertical rupture front distorts the pulse shape, but does not have an important effect on peak amplitudes. One may interpret the motion caused by a finite fault, with constant rupture velocity and slip, rupturing past a site as consisting of a starting phase, a phase associated with the passage of the rupture event, and a stopping phase. The motions caused by the infinite-length fault dislocation model, considered in this paper, correspond to the rupture passage phase. Compared with empirical correlations and observations in Imperial Valley, California, the model yields peak amplitudes of acceleration which are typically a factor of 6 too small. Effects caused by layered structure in the earth, which are absent from the dislocation model, are the most likely source of this discrepancy.