Not long ago the sample of well studied supernovae, which were gathered mostly through targeted surveys, was populated exclusively by events with absolute peak magnitudes fainter than about -20. Modern searches that select supernovae not just from massive hosts but from dwarfs as well have produced a new census with a surprising difference: a significant percentage of supernovae found in these flux limited surveys peak at -21 magnitude or brighter. The energy emitted by these superluminous supernovae in optical light alone rivals the total explosion energy available to typical core collapse supernovae (> 1051 erg). This makes superluminous supernovae difficult to explain through standard models. Adding further complexity to this picture are the distinct observational properties of various superluminous supernovae. Some may be powered in part by interactions with a hydrogen-rich, circumstellar material but others appear to lack hydrogen altogether. Some appear to be powered by large stores of radioactive material, while others fade quickly and have stringent limits on 56-Ni production. In this talk I will discuss the current observational constrains on superluminous supernova and the prospects for revealing their origins.