Repeated neuropsychological assessments are common with older adults, and the determination of true neurocognitive change is important for diagnostic assessment. Several statistical formulas are available to assist in this determination, but they rely on access to test-retest stability coefficients and practice effect values. The current study presents data on these psychometric properties of the RBANS in a large community dwelling elderly sample. Across a one-year retest interval, stability coefficients ranged from .58 to .83 for the Index scores, and from .51 to .83 for the subtest scores. Practice effects were largely absent, with most performances slightly decreasing at retest. These psychometric properties are contrasted with those reported in the RBANS manual, and possible reasons for these differences are discussed. A case example is provided that demonstrates the use of the current findings in conjunction with existing change formulas.