The prevalence of autism spectrum disorder (ASD) has risen rapidly in the past decade. Estimates on factors associated with developmental screening and the timing of events along the diagnostic pathway can inform early identification efforts. This dissertation uses cross-sectional data from the 2016–2018 National Survey of Children’s Health to achieve 3 aims: 1) evaluate individual- and state-level sources of variance between states in developmental screening rates via multilevel models, 2) evaluate characteristics associated with the ages at which children with ASD are first diagnosed, receive an intervention plan, and begin intervention, and 3) evaluate differences in lengths of time between these events by cohort. Aim 1: The national rate of developmental screening for children ages 9 months to 5 years is 34.4% (95% Confidence Interval (CI), [34.3, 34.4]). Rates varied between states by 38%. Individual-level factors explained 6% of the variance, while income inequality and a state’s choice to track developmental screening did not explain any variance between states. Aim 2: Linear regression models adjusted for individual and household characteristics showed that compared to children aged 3–5 years at the time of the survey, children 6–11 were 18 months older at first services (? =1.49, 95% CI, [1.18, 1.81] and children aged 12–17 were 38 months older at first ASD diagnosis (? =3.16, 95% CI, [2.72, 3.60]. Aim 3: Analyses using identical models showed that compared to children aged 3–5 at the time of the survey, the interval between first plan and first services was 4 months longer for children 6–11 (? =0.34, 95% CI, [0.07, 0.61]; and 8 months longer between first ASD diagnosis and first services for children aged 12–17 (? =0.67, 95% CI, [0.28, 1.06]. Today’s children with autism receive their first diagnosis, intervention plans, and developmental services at younger ages than in the past and are moving between events with less delay compared to older children. However, the low rate of developmental screening nationwide represents missed opportunities for even earlier identification. Research is needed to identify the macro-level factors that explain the variance between states on developmental screening rates.