This paper describes and experimentally demonstrates the main tenets of an operant theory of leadership. Leadership is characterized in the current paper as involving problem solving operant behavior (Cerutti, 1989; Skinner, 1969) in a social context (Skinner, 1953). The theory was assessed under two experimental analogs modeled from generic formal organizational bureaucratic leader-follower role relations. Under a minimal leadership contingency (MLC) leaders and followers in N = 4 dyads interacted via button pressing and trigger pulling responses, respectively, and they received feedback on counters located on response panels in their separate rooms. Under the MLC every leader button press added a point worth money to one of the follower's counters but the leader received no points worth money based on follower responses. A leadership contingency (LC) was identical to the MLC except that for every 19th follower trigger pull the leader received a point worth money. As anticipated, high rates of leader-follower interaction evolved in all dyads under the LC and appreciably lower rates occurred under the MLC as leader button pressing extinguished under the MLC with repeated exposures to the two contingencies presented in ABABAB fashion. Results were discussed in terms of the theory and data as they may be related to assessment and maintenance of leader-follower interactions and performance in OBM lab and field experiments.