Increasing accumulation and retention of nanomedicines within tumor tissue is a significant challenge, particularly in the case of brain tumors where access to the tumor through the vasculature is restricted by the blood–brain barrier (BBB). This makes the application of nanomedicines in neuro-oncology often considered unfeasible, with efficacy limited to regions of significant disease progression and compromised BBB. However, little is understood about how the evolving tumor–brain physiology during disease progression affects the permeability and retention of designer nanomedicines. We report here the development of a modular nanomedicine platform that, when used in conjunction with a unique model of how tumorigenesis affects BBB integrity, allows investigation of how nanomaterial properties affect uptake and retention in brain tissue. By combining different in vivo longitudinal imaging techniques (including positron emission tomography and magnetic resonance imaging), we have evaluated the retention of nanomedicines with predefined physicochemical properties (size and surface functionality) and established a relationship between structure and tissue accumulation as a function of a new parameter that measures BBB leakiness; this offers significant advancements in our ability to relate tumor accumulation of nanomedicines to more physiologically relevant parameters. Our data show that accumulation of nanomedicines in brain tumor tissue is better correlated with the leakiness of the BBB than actual tumor volume. This was evaluated by establishing brain tumors using a spontaneous and endogenously derived glioblastoma model providing a unique opportunity to assess these parameters individually and compare the results across multiple mice. We also quantitatively demonstrate that smaller nanomedicines (20 nm) can indeed cross the BBB and accumulate in tumors at earlier stages of the disease than larger analogues, therefore opening the possibility of developing patient-specific nanoparticle treatment interventions in earlier stages of the disease. Importantly, these results provide a more predictive approach for designing efficacious personalized nanomedicines based on a particular patient’s condition.