The accuracy of the percentage of survival technique (P-S Method) for predicting school enrollment is examined by testing it on 242 towns and cities in Massachusetts. An attempt is also made to identify certain factors associated with accuracies and inaccuracies in the P-S Method. Three general conclusions are drawn: (1) The P-S Method produces the greatest error in fast-growing communities that most need accurate enrollment forecasts, (2) enrollment change is a fairly good predictor of forecast accuracy, and (3) in larger school systems a forecast that is technically "accurate" (within 10 percent error) may be inaccurate by several hundred or even several thousand students. Four general classifications of factors affecting population and thus enrollment are (1) The number of dwellings available in the community, (2) the type of people living in those dwellings, (3) the demand for dwellings in the community, and (4) factors controlled by public and nonpublic schools. A number of sources of information can be tapped by the school administrator to allow him to make better enrollment forecasts. (HW)