The relationship between blended, virtual, and traditional learning on student performance can significantly affect the future of physical education programs’ curriculum and instructional methods. This causal-comparative research design study examines the difference in student performance measured by FitnessGram® scores between ninth-grade female physical education students who participate in an all-online/virtual learning instructional model, a blended learning instructional model, and a traditional face-to-face learning instructional model. Using a random sample of 143 students from six physical education classes from two high schools in a rural South Carolina school district, a one-way ANOVA was conducted to examine the difference in student performance measured by the FitnessGram® based on learning models. The results were not statistically significant, suggesting that there is no significant difference in student performance as measured by FitnessGram® scores among female ninth-grade physical education students who participate in an all-online/virtual learning instructional model, a blended learning instructional model, and a traditional face-to-face learning instructional model. Recommendations for future research will suggest a larger sample size, including male students, and further investigation identifying which student population benefits most from virtual physical education.