Development of the neurovascular unit (NVU) is a complex, multistage process that requires orchestrated cell signaling mechanisms across several cell types and ultimately results in formation of the blood-brain barrier. Typical high-throughput screening (HTS) assays investigate single biochemical or single cell responses following chemical insult. As the NVU comprises multiple cell types interacting at various stages of development, a methodology combining high-throughput results across pertinent cell-based assays is needed to investigate potential chemical-induced disruption to the development of this complex cell system. To this end, we implemented a novel method for screening putative NVU disruptors across diverse assay platforms to predict chemical perturbation of the developing NVU. HTS assay results measuring chemical-induced perturbations to cellular key events across angiogenic and neurogenic outcomes in vitro were combined to create a cell-based prioritization of NVU hazard. Chemicals were grouped according to similar modes of action to train a logistic regression literature model on a training set of 38 chemicals. This model utilizes the chemical-specific pairwise mutual information score for PubMed MeSH annotations to represent a quantitative measure of previously published results. Taken together, this study presents a methodology to investigate NVU developmental hazard using cell-based HTS assays and literature evidence to prioritize screening of putative NVU disruptors towards a knowledge-driven characterization of neurovascular developmental toxicity. The results from these screening efforts demonstrate that chemicals representing a range of putative vascular disrupting compound (pVDC) scores can also produce effects on neurogenic outcomes and characterizes possible modes of action for disrupting the developing NVU.