We use experimental data from the Head Start Impact Study to examine the effect of sequential participation in childcare programs on cognitive outcomes. Using a sequential threshold model, which accounts for selection beyond initial randomization, we estimate causal returns to program sequences, including joint and cross returns across skill investment programs. We then estimate a dynamic structural model as a juxtaposition and evaluate a counterfactual policy reform which limits individuals to one year of Head Start. Our results support Head Start implementation earlier in life, and support engaging low-ability children with center-based care and high-ability children with some home care.