There are many situations in which human performances appear insensitive to changing contingencies of reinforcement when compared to nonhuman operant performances. Explanations of these discrepancies have appealed to rule-governance and have provided some evidence that instructions produce these differences by restricting response alternatives as well as functioning as discriminative stimuli for other contingencies. In order to further evaluate these potential functions, a canonical study on rule-governance was systematically replicated. Five undergraduate participants were tasked with earning blocks by pressing a button during LED-signaled, fixed-ratio 7 and differential-reinforcement-of-low-rate 5-s schedules of reinforcement. Phase 1 of the experiment switched between these two schedules, with the schedule alternating every 1 minute. Phase 2 added instructions to "Go Fast" and "Go Slow" to the LEDs and programmed the lit LED to switch 30 seconds into each 1-minute session. Phase 3 removed the instructions from the LEDs and returned to the procedures of phase 1, with only one LED lit during each 1-minute session. Results showed that instructions influence the response rates as well as stimulus control over those rates. Results also showed that all participants ignored instructions conflicting with the reinforceable rate by the end of Phase 2. These findings indicate that instances of insensitivity may result from instructions restricting response alternatives and their stimulus control, as opposed to instructions signaling a second contingency. These findings are discussed in terms of the nature of instructional stimulus control, its relation to the operant unit of analysis, and alternative interpretations of other instances of insensitivity.