MRI estimations of cerebral blood volume (CBV), useful in mapping brain dysfunction, typically require intravenous (IV) injections of contrast agents. Transgenically engineered mice have emerged as the dominant animal model with which to investigate disorders of the brain and novel therapeutic agents. The difficulty in gaining IV access in mice prohibits repeated administration of contrast in the same animal, limiting the ability to map CBV changes over time. Here we address this limitation by first optimizing an approach for estimating CBV that relies on intraperitoneal (IP) rather than IV injections of the contrast agent gadodiamide. Next, we show that CBV maps generated with IP or IV injections are quantitatively comparable. Finally, we show that CBV maps generated with IP gadodiamide can be acquired repeatedly, reliably and safely over time. Although this approach has certain limitations, estimating CBV with IP injections is well-suited for mapping the spatiotemporal pattern of brain dysfunction in mice models of disease, and for testing pharmacological agents.