The main goal of this paper is to test the shirking version of the efficiency wage model for the Brazilian industries' case. This model states that there exists a negative relationship between wages and intensity of supervision, or a tradeoff between internal and external supervision. Different from other empirical papers developed for the Brazilian case, this paper uses the span of control (supervision/ staff ratio) - rather than size of firms - as proxy variable for intensity of supervision. The efficiency wage hypothesis is empirically supported by this work.