Traveling interplanetary shocks can accelerate protons up to energies of some tens of MeV. Intensity and anisotropy time profiles observed in interplanetary space can be used to determine the scattering conditions as well as the evolution of the shock efficiency along the observer's magnetic field line. The latter can be described as a superposition of an azimuthal variation along the shock front, which accounts for the dependence of intensity time profiles on the angular distance between the observer and the nose of the shock, and a radial variation, which probably is related to the energetics of the solar event and the shock. Fits with a black box model to 44 particle events lead to the following results: (1) overall scattering conditions in interplanetary space are not different in events with and without interplanetary shock, (2) the radial variation S~r&agr; of the injection has a power law index &agr; between -5.5 and +4.5 with &agr;≤0 for most of the events, (3) a separate solar component is required in four of 44 events only, (4) within the broad scatter from event to event, &agr; and &lgr; are independent of whether the observer is located east or west of the flare normal and (5) if the same shock has been observed by two spacecraft, observations on both spacecraft can be described consistently. Results (4) and (5) suggest that the description of the injection in the black box model is a reasonable approximation. The other results give hints on the acceleration: First, there appears to be no significant enhancement in overall interplanetary turbulence which could lead to a more efficient acceleration. Second, in most of the events presented here the number of injected particles per unit surface of the shock is highest close to the Sun; in about half of the events also the number of particles expressed as fraction of the ambient medium is highest close to the Sun. This has two implications: (1) either there is a strong contribution of flare-accelerated particles or, in the context of current knowledge more likely, the shock is a more efficient particle accelerator close to the Sun than in interplanetary space, and (2) the relatively small amount of particles accelerated locally in interplanetary space does not require significant acceleration, it might even be reaccelerated material from the large population of particles accelerated close to the Sun. ¿ 1997 American Geophysical Union