Probe tests provide an important additional converging operation to identify the characteristics of cognitive processes being assessed by other behavioral measures. We introduce a new procedure using a variable-interval (VI) probe test to assess spatial memory in the water maze. Three groups of rats were trained in spatial discrimination with traditional platform trials. For the probe trials, 2 groups had no-platform (NP) probe trials in which the escape platform was unavailable for the entire trial. The 3rd group had variable interval (VI) probe trials, in which the escape platform was available only after a VI. The VI probe trial was a more sensitive measure of spatial memory than was the NP probe trial and was useful for repeated measurements of spatial memory, a characteristic not readily attainable with the NP probe trial.