Using three-dimensional diffusion-weighted MR microscopy and a rat model of focal cerebral ischemia, we evaluated the statistical characteristics of two parameters: absolute stroke volumes and change in stroke volumes over 6 h of middle cerebral artery (MCA) occlusion. In all rats, the absolute stroke volumes increased linearly over the 6-h MCA occlusion time period. On average, stroke volume growth rate was 2.1 +/- 0.5%/h. Sample size power analysis of our data demonstrated that to demonstrate a 10% reduction of the 6-h volumes, sample size per group would require 29 animals (these calculations are based on alpha = 0.05, beta = 0.20 using normal approximation). A similar 30% reduction of stroke volume at 6 h poststroke (from approximately equal to 200 to 140 mm3) would, in our "slope model," translate into a reduction of stroke growth rate from the normal + 11.25 mm3/h (150 to 200 mm3 over 4 h) to 7 mm3/h (150 to 178 mm3 over 4 h); power analysis in this case demonstrated that sample size is reduced to 15 animals per group (these calculations are based on alpha = 0.05, beta = 0.20 using normal approximation). We conclude that from a statistical standpoint our study demonstrates that stroke growth rate might be a more suitable parameter for evaluating the effect of treatment in both clinical and experimental stroke trials.