Background: The long-term effectiveness of Head Start remains a topic of debate. Previous research with families of children enrolling in Head Start at three years of age has shown patterns of familial risk moderate the short-term effects of Head Start suggesting some children benefit more than others. However, we know little about patterns of risk that moderate Head Start's effectiveness for children who first enroll in Head Start at four-year-old and even less about risk factors that moderate the long-term effectiveness of Head Start. Objective: In the current study, we sought to address these gaps by examining differential treatment effects of Head Start on elementary school outcomes for subgroups of children who first enrolled in Head Start at age four. Methods: Latent class analysis (LCA) was used to identify subgroups of children from the four-year-old cohort of the Head Start Impact Study (N = 2,108) based on family risk variables in preschool (e.g., maternal depression). Weighted regression analyses were then conducted to assess if the effects of Head Start on end of pre-kindergarten, kindergarten, 1st, and 3rd grade academic and social-emotional outcomes varied by latent class membership. Results: While a subsample of children attending Head Start experienced sustained benefits, others had worse outcomes in elementary school, and predominantly any initial benefits of Head Start were not sustained into elementary school. Conclusion: Irrespective of patterns of familial risk, initial benefits gained via Head Start for children first enrolling in at four-years of age typically fade out during elementary school.