Abstract: Data extraction for semiconductor process analysis may be implemented across multiple databases. A user may select a given level of interest, such as a wafer, a lot, or a die, and may extract specified information across more than one database if desired. The databases may include separate information such as process control information, electrical, and sort test information. Instead of coalescing the databases into one extremely unmanageable database, data can be extracted horizontally across those databases using structured queries.