OBJECTIVES--To compare the pullout properties of 3.5-mm AO/ASIF self-tapping screws (STS) to corresponding standard cortex screws (CS) in a uniform synthetic test material and in canine femoral bone. The influence of screw-insertion technique, test material, and test-material thickness were also assessed. STUDY DESIGN--In vitro experimental study. SAMPLE POPULATION--Two independent studies: a uniform synthetic test material and paired femurs from mature dogs. METHODS-Mechanical testing was performed in accordance with standards established by the American Society for Testing and Materials for determination of axial pullout strength of medical bone screws. Completely inserted STS, completely inserted CS, and incompletely inserted STS were tested in 3 groups of 10 test specimens each in 4.96-mm and 6.8-mm thick sheets of synthetic material. In the bone study, group 1 consisted of 24 completely inserted STS compared with 24 completely inserted CS, and group 2 consisted of 24 incompletely inserted STS versus 24 completely inserted CS. Comparisons were made between paired femurs at corresponding insertion sites. Pullout data were normalized, thereby eliminating the effect of test-material thickness on pullout properties. Mean values were compared using 2-way ANOVA. Statistical significance was set at P <.05. RESULTS--In both the 4.96-mm and 6.8-mm synthetic material, pullout testing of the completely inserted STS demonstrated significantly greater yield strength and ultimate strength than completely inserted CS. There was no significant difference between incompletely inserted STS and completely inserted STS. The 6.8-mm test material significantly increased yield strength and ultimate strength for all test groups compared with the 4.96-mm test material. In canine bone, there was no significant difference in yield strength of completely inserted STS and completely inserted CS. Yield strength of completely inserted STS and completely inserted CS were significantly greater than incompletely inserted STS. CONCLUSIONS--Pullout properties of completely inserted STS were significantly greater than corresponding CS in a uniform test material. In canine bone, the pullout strength of STS and CS were not different. Incomplete STS insertion resulted in an 18% reduction in holding power as compared with completely inserted CS and STS in canine bone. CLINICAL RELEVANCE--The length of STS used in canine bone should be such that the cutting flutes extend beyond the trans cortex to maximize pullout strength.