Previous studies have shown that low-efficacy benzodiazepines may function as full agonists, partial agonists or antagonists, depending upon the sensitivity of the assay to detect a drug's agonist effects. To date, these differential effects have only been observed across tasks, as these drugs rarely produce full agonist and antagonist effects in a single preparation. The purpose of the present study was to examine the agonist and antagonist effects of various benzodiazepines in a motor task in which the sensitivity of the task differed across conditions. To this end, rats were trained to walk on a rotorod apparatus rotating at low (4 rpm), medium (20 rpm) and high (36 rpm) speeds, and the motor-impairing effects of representative low (bretazenil), intermediate (clonazepam) and high (diazepam) efficacy benzodiazepines were examined. Diazepam and clonazepam reduced the latency for rats to fall from the rotorod at all three speeds, with both drugs being more potent at the high speed than at lower speeds. Bretazenil was also effective at decreasing motor performance at the high speed, but was not effective at lower speeds. In drug combination tests, bretazenil antagonized the effects of diazepam and clonazepam under conditions in which it was not effective when administered alone, effectively demonstrating both full agonist and antagonist effects in the same preparation. These data indicate that the sensitivity of a motor task to detect a drug's agonist effects can be manipulated by altering its relative level of difficulty, and that lower-efficacy benzodiazepines are more sensitive than higher-efficacy benzodiazepines to these manipulations.