Control phenomena have been at the center of syntactic theorizing almost since the inception of generative grammar. To first approximation, control is an interpretive dependency between two arguments: a lexicalized noun phrase (the controller) and a null argument in a dependent clause (the controllee), as in
Sandy
hopes [
∅
to succeed]
. The primary concerns of syntactic theories have been accounting for various characteristics of the null controllee and the identification of its controller. This chapter surveys some of the phenomena that have driven our current understanding of these issues. Within the domain of obligatory control, some phenomena speak to the controller: the behavior of polyadic predicates, implicit control, variable control and control shift, and exceptional cases like
promise
. Other phenomena speak to the controllee: exhaustive control, partial control, split control, and control in embedded questions. Assuming that non‐obligatory control is a kind of elsewhere case, the phenomena relevant to non‐obligatory control are those where obligatory control restrictions on controller choice fail to materialize: control into subjects, extraposed clauses, and some adjunct clauses. The chapter also discusses phenomena with a non‐canonical realization of the control relation: backward control and copy control. Within the context of these phenomena, the chapter surveys current generative approaches to the analysis of control: dominant PRO‐based approaches, the movement theory of control, and restructuring analyses. Relative strengths and weaknesses are highlighted.