When multiple regression is used to develop a prediction model, sample size must be large enough to ensure stable coefficients. If sample size is inadequate, the model may not predict well in future samples. Unfortunately, there are problems and contradictions among the various sample size methods in regression. For example, how does one reconcile differences between a 15:1 subject-to-variable ratio and a 30:1 rule. The purpose of this study was to validate a precision power method for determining sample sizes in regression. The method uses a cross-validity approach to selecting sample sizes so that models will predict as well as possible in future samples. The simple formula, which is an algebraic manipulation of a cross-validation formula, enables researchers to limit the expected shrinkage of R squared. Using a Monte Carlo simulation study, the precision power method was compared to eight other methods. It was the only method that provided consistently accurate and acceptable precision power rates. That is, when precision power was set a priori, actual precision power rates consistently fell within an acceptable interval around that given power rate. (Contains 3 tables and 78 references.) (Author/SLD)