Performance errors committed during the acquisition of a carrier landing task in the simulator resulted in the occurrence of a freeze. Pilot subjects exposed to the freeze developed control strategies which were distinguishable from those of a no-freeze control group in terms of throttle, rudder, aileron, and elevator activity. Neither rate of learning nor level of performance was affected. Use of the freeze, however, was reported as initially "frustrating" and as adding to the overall difficulty of acquiring the task. With certain applications, the freeze appears to be aversive, altering pilot motivation for learning the task. Methodologically, the study's use of a probe trial technique represents a departure from the traditional transfer of training methodology for the assessment of alternative instructional treatments.