Since it has been an obligation to submit microarray experiment data to public repositories for publishing the experimental results, collecting and storing the data is an essential step in microarray study. Implemented with H2 DBMS as the backend storage, Magnolia provides PFGRC array users with a series of intelligent entry form to collect the information or parameters of experiment, bio-source, sample, probe, slide, scan, expression, and normalization, which is MIAME compliant (Minimal Information About a Microarray Experiment; (Brazma et al., 2001)). The collected data are saved to a relational database defined by users and can be modified easily with a set of corresponding edit pages. Working as an exporter, Magnolia can write out the entered data into GEO sample and series files in SOFT format that is acceptable by GEO, so that the files can be submitted to GEO directly. Magnolia also serves as a platform for launching other data manipulation tools, such as SQL Query Window to report the entered data, Slide Annotation to load array annotation file downloaded from PFGRC site into the database, and Database Schema to present the relationships among the tables. Registration not required.