In this talk, as requested, I begin with a overview and with some basic reminders about how evidence for supersymmetry in nature might appear - in particular, how SUSY signatures are never clear so it is difficult to search for them without major theoretical input. Models can be usefully categorized phenomenologically by naming their LSP - that is, once the LSP is approximately fixed so is the behavior of the observables, and the resulting behavior is generally very different for different LSPs. Next I compare the three main LSP-models (gravitino, bino, higgsino). Hints from data suggest taking the higgsino-LSP world very seriously, so I focus on it, and describe its successful prediction of reported events frost the 1990 LEP runs. SUSY signatures in the (h) over tilde LSP world are very different from those that are usually studied. Then I briefly discuss how to measure the parameters of the effective Lagrangian from collider and decay data. Finally I turn to how data will test and help extract the implications of string theories.