A methodology is described for use of a 16-hole circular platform task to test spatial memory in mice. Both bright light and a fan were used to motivate mice to escape the platform surface through a single hole containing an attached escape box. For each daily trial, three correlated measures (escape latency, number of errors, and error rating) comprehensively evaluated cognitive performance. In an initial study, the 'spatial' nature of this task was demonstrated by the much poorer performance of non-transgenic mice when visual cues are removed. Behavioral sensitivity of the circular platform task was then shown through its ability to discern cognitive impairment in 7-month-old transgenic mice, carrying the mutant APP(SW) gene for early-onset Alzheimer's disease in humans, from non-transgenic litter-mates. Since there are currently only a few tasks available to definitively test cognitive performance in mice, the circular platform task offers a versatile, multiple-measure option with numerous advantages. Particularly in view of the increasing number of genetically manipulated mouse models being produced, the circular platform task should be most useful in providing a sensitive evaluation of cognition in mice.