Growing evidence indicates that early motor development and motor behavior differ between autistic and neurotypical children. This has led to increased interest in the utility of motor behavior for detecting and understanding the unfolding development of autism spectrum disorder (ASD). Differences in gait and independent walking may be particularly relevant to the early development of ASD. Atypical gait features have not only been proposed as a predictive biomarker of ASD, but have been hypothesized to negatively impact autistic children’s communication development by altering their engagement with the world around them. However, no studies to date have assessed early gait in infants who go on to have ASD. The present dissertation addresses this gap by conducting quantitative gait assessments in infants with and without a family history of ASD approximately one month after they achieved independent walking and ascertaining their diagnostic outcomes at two years of age. Chapter 1 examined the relationship between early gait and language development, specifically testing whether differences in basic motor control predicted change in the rate of language acquisition following walk onset. Results demonstrated that although autistic children acquired language more slowly than neurotypical children overall, rates of language acquisition significantly increased following walk onset regardless of diagnostic outcome. Furthermore, gait control was associated with the overall rate of growth in children’s receptive vocabulary, but was not found to mediate the relationship between walk onset and increased language development as predicted. Chapter 2 investigated whether early gait features predicted later ASD and language outcomes. Greater gait variability was significantly and specifically associated with more ASD symptoms at 24 months of age. Moreover, machine learning models based on three-dimensional (3D) video-based gait features classified later ASD diagnostic outcomes with higher positive (49%) and negative predictive values (89%) than an established parent-report screening measure. To support accessibility and scalability of early gait assessment, Chapter 3 leveraged markerless motion capture techniques to develop and validate a new 3D video-based infant gait assessment tool. Taken together, the present findings highlight early gait as a meaningful marker of neurodevelopment and as a useful, scalable tool for enhancing early detection of ASD.