Autism shows a pronounced and replicable sex bias with approximately three-to-four males diagnosed for every one female. Sex-related biology is thought to play a role in the sex bias, such that female biology may be protective and/or male biology may increase vulnerability to autism in the context of similar genetic risk. Beyond etiology, sex-related biology has also been implicated in lifespan risk for health and psychiatric conditions that show common co-morbidity in autism. Thus, understanding how sex-related biology impacts autism etiology and progression has important implications for prognosis and treatment. Neuroimaging offers a powerful tool for in-vivo characterization of brain-based sex differences in autism, especially given emerging efforts to develop large, well-characterized longitudinal samples. To date, however, neuroimaging studies have shown mixed and inconsistent findings, which remain challenging to integrate in the broader literature context. In a recent systematic review of neuroimaging studies of typical sex differences, few to no replicable effects were found beyond brain size, suggesting the brain is not “sexually dimorphic.” Instead, it is argued that the brain is a “mosaic” of features from various sources, including masculine and feminine biological processes as well as individual genetics and environment. Thus, designing neuroimaging studies that are sensitive to brain-based sex differences in autism likely requires careful study design and analytical method selection. Through a series of studies, the overarching dissertation aim was to identify optimal methods for characterizing neuroimaging-based sex differences in autism and to test these methods in preliminary samples. Study 1 comprised a systematic review of studies examining neuroimaging-based sex differences in autism with the aim of identifying optimal study designs, neuroimaging modalities, and analytical methods. Study 2 focused on examining the sensitivity of a connectome-wide approach to identify functional connectivity hubs underlying sex-biased behavior associated with autism (e.g., camouflaging). Study 3 used a connectome-wide functional connectivity approach to characterize sex differences in longitudinal changes associated with autistic traits vs. categorical diagnosis. These studies suggest that optimizing study design and methods improves identification of biologically plausible and clinically meaningful brain sex differences in autism. The relevance of findings to etiology and prognosis are discussed.