The purpose of the present experiment was to examine the effectiveness of a modified rotarod test in detecting motor deficits following mild and moderate central fluid percussion brain injury. In addition, this investigation compared the performance of the rotarod task with two other commonly used measures of motor function after brain injury (beam-balance and beam-walking latencies). Rats were either injured with a mild (n = 14) or moderate (n = 8) level of fluid percussion injury or were surgically prepared but not injured (n = 8). All rats were assessed on all tasks for 5 days following their respective treatments. Results revealed that both the mild and moderate injury levels produced significant deficits in the ability of the animals to perform the rotarod task. Performance on the beam-balance and beam-walking tasks were not significantly impaired at the mild injury level. It was only at the moderate injury level that the beam-balance and beam-walking tasks detected deficits in motor performance. This result demonstrated that the rotarod task was a sensitive index of injury-induced motor dysfunction following even mild fluid percussion injury. A power analysis of the three tasks indicated that statistically significant group differences could be obtained with the rotarod task with much smaller sample sizes than with the beam-balance and beam-walking tasks. Performance on the rotarod, beam-walk, and beam-balance tasks were compared and evaluated by a multivariate stepdown analysis (multiple analysis of variance followed by univariate analyses of covariance). This analysis indicated that the rotarod task measures aspects of motor impairment that are not assessed by either the beam-balance or beam-walking latency. These findings suggest that compared to the beam-balance and beam-walking tasks, the rotarod task is a more sensitive and efficient index for assessing motor impairment produced by brain injury.