Stellar multiplicity is a ubiquitous outcome of the star-formation process. The frequency and main characteristics of multiple systems, and their dependence on primary mass and environment, are powerful tools to probe this process. Although early attempts were fraught with selection biases and limited completeness, instrumentation breakthroughs in the past two decades now enable robust statistical analyses. In this review, we summarize current empirical knowledge of stellar multiplicity for main sequence stars and brown dwarfs, as well as among populations of pre-main-sequence stars and embedded protostars. Among field objects, the multiplicity rate and breadth of the orbital period distribution are steep functions of the primary mass, whereas the mass ratio distribution is essentially flat for most populations other than the lowest mass objects. The time-variation of the frequency of visual companions follows two parallel, constant tracks corresponding to loose and dense stellar populations, although current observations do not yet distinguish whether initial multiplicity properties are universal or dependent on the physical conditions of the parent cloud. Nonetheless, these quantitative trends provide a rich comparison basis for numerical and analytical models of star formation.