The significant etiologic and phenotypic heterogeneity of autism spectrum disorder (ASD) has made it challenging to target underlying mechanisms of ASD pathology and identify replicable biomarkers. A burgeoning “genetics-first” approach has been proposed to reduce heterogeneity and enable identification of subtypes of individuals with ASD across multiple physiological and behavioral systems. To make progress in the development of ASD biomarkers, we must establish the connection between behavior and the underlying physiology associated with etiologically defined subgroups. In this panel, we will focus on recent discoveries spanning multiple units of analysis to describe ASD phenotypes of children and individuals with genetic variants. We will begin by identifying behavioral patterns of syndromic ASD associated with de novo mutations and copy number variations (Stessman). We will then present data connecting behavior and physiology for specific genetic subtypes, including eye gaze phenotypes in PTEN carriers (Frazier), EEG resting state phenotypes in 15q11.2-13 duplication carriers (Jeste), and EEG dynamic patterns of attention in SCN2A carriers (Hudac). We will discuss common strategies and underlying mechanisms across these topics and the relevance of genetically distinct subtypes for the identification of biomarkers in ASD.