Public data resources increase the impact of experimental neuroscience, enabling students, neural data scientists, and experimentalists to learn from and produce new findings out of a single data collection. The Allen Brain Observatory resource makes public large collections of optical (“ophys”) and electrical (“ephys”) neuronal recordings obtained from multiple brain areas in awake, behaving mice during visual stimulation. Using the Allen Brain Observatory’s RESTful API, the Brain Observatory Toolbox (BOT) provides an interactive, high-level object-oriented interface within MATLAB to query, inspect, and access this data. The data is organized into dataset items: experimental sessions, Neuropixels recording probes, and their channels. Derived dataset items such as cells (for ophys) and units (for ephys) are also provided. Users can fetch tabular arrays of available dataset items and apply intuitive tabular indexing semantics to inspect and select specific items for further study. For example, users could selectively filter out sessions from specific brain areas, transgenic (Cre) lines, and stimulus conditions. Object definitions are also provided for each dataset item type and can be constructed from unique item identifiers or row selection semantics. These allow inspection of data item information, implement local caching of data files, and provide a command-line interface of properties and methods to extract further information from the data files of Neurodata without Borders (NWB) and other filetypes. The Brain Observatory Toolbox is now available as an early-version (0.9) GitHub release. User feedback is welcomed, and further coverage of newer Allen Brain Observatory resource contents is planned.