Data users in government, private industry, non-profits and academia have substantial demand for data from the Census Bureau’s censuses and surveys. Hence, the Census Bureau aims to disseminate data widely and with as much detail as possible while keeping the pledge of confidentiality given to all respondents. Although the Census Bureau produces many estimates, tables, and public use microdata , existing data products don’t always meet users’ needs. This paper describes the development and capabilities of the Microdata Analysis System (MAS), an online remote access system in development at the Census Bureau. The MAS will allow any user to request custom tables and analyses from the underlying microdata. Tables and estimates will be produced quickly and interactively for the user’s specified geography and analysis variables. The system will also dynamically generate measures of variance for all estimates. Data will be protected to ensure respondent confidentiality. We describe the capabilities of the system and evaluate the planned disclosure avoidance methodology to protect the data and the system’s data utility. We also outline challenges with the planned methodology for future research.