We analyzed the early spica casting outcomes of 50 children age 2 to 10 years with uncomplicated femoral shaft fractures treated at Johns Hopkins Hospital between October 1987 and October 1990. Our objective was to develop criteria for the prospective identification of patients who can be safely and dependably treated with early spica casting without excessive shortening of the fracture fragments. Forty-one (82%) children had an acceptable outcome and nine (18%) had an unacceptable outcome according to our definition of > 25 mm of fracture fragment overlap at 3 to 4 weeks follow-up. A new clinical test, the telescope test, was statistically significant (p < 0.001) for association with spica casting outcome. Age, sex, fracture location, mechanism of injury, fracture type, and resting roentgenogram fracture fragment overlap were not statistically significant (p > 0.10). The telescope test had a sensitivity of 78%, a specificity of 85%, and a negative predictive value of 95% for predicting spica casting outcome. The relative risk of failing spica casting after a positive telescope test was 20.4 (95% CI, 2.74-225.10). We conclude that children 2 to 10 years of age with uncomplicated femoral shaft fractures and a negative telescope test can be safely treated with early spica casting and have a 95% change of having a successful outcome with this treatment.