This experiment explored the distribution of attention in a word learning situation among three groups of children: typically developing (TD group), language delayed (LD group), and children with autism spectrum disorders (AD group). Participation involved two visits: a screening visit conducted in a developmental clinic and an eye tracking visit conducted in the laboratory. At the screening visit, diagnostic, cognitive, and language testing were administered and parents completed surveys. At the eye tracking visit, children’s fixations were monitored while they watched videos of a modified version of a word-learning paradigm used by Baldwin (1991). Order of the visits was based on the schedule of the family. Parents provided informed consent at the child’s first visit. Power analyses were run (Lenth, 2006-2009) in order to confirm that the current sample size was sufficient to test for a significant effect given effect size estimates for differences in attention to social stimuli obtained in previous studies (Chawarska et al., 2012).