Mechanisms such as grain rearrangement, coupled with elastic deformation and grain breakage, are believed to play an important role in the time-independent compaction of sands, controlling porosity and permeability reduction during burial of clastic sediments and during depletion of highly porous reservoir sandstones. We performed uniaxial compaction experiments on sands at room temperature to systematically investigate the effect of loading history, loading rate, grain size, initial porosity, and chemical environment on compaction. Acoustic emission counting and microstructural methods were used to verify the microphysical compaction mechanisms operating. All tests showed quasi-elastic loading behavior accompanied by permanent deformation, involving elastic grain contact distortion, particle rearrangement, and grain failure. Loading history, grain size, and initial porosity significantly affected stress-strain behavior, with increasing grain size and initial porosity promoting compaction. In contrast, chemical environment and loading rate had little effect. The results formed the basis for a microphysical model aimed at explaining the observed compaction behavior. Two extreme cases were modeled: (I) a pack of spherical grains with a distributed flaw size at failure and (II) a pack of nonspherical grains with a constant mean crack size at failure but a distributed effective surface radius of curvature characterizing distributed contact asperity amplitudes. The best agreement with the grain-size-and porosity-dependent trends observed in our experiments was obtained using case (II) of the model. Combining our experimental and modeling results, it was inferred that a grain-size-dependent departure from sphericity of the grains exerts a key control on the compaction behavior of sands.