The Open Microscopy Environment (OME) is a framework for the management and analysis of image data and metadata in biological microscopy. Biological microscopy images can be collected in many different ways, and may represent many different kinds of information which changes continually with evolving technology and experimental goals. A framework that fully encompasses biological microscopy in scope cannot rely on a fixed data model it must be designed to accommodate ever-changing informatics needs. The challenge posed by a fluid data model is similar to one addressed by the semantic web, principally support for arbitrary or user-defined semantics and ontologies. However, an analysis framework faces additional challenges besides the management of semantics: definition of analytic units as transforms between semantic constructs, definition of aggregates of analytic units to represent work flows, interfaces to algorithm implementations, and maintenance of data provenance or history. A collaborative scientific environment also demands that these locally defined semantics and transforms be fully transportable for subsequent review or analysis. OME incorporates these components in a database-backed system targeted at end-user biologists. This presentation will briefly describe the various components of OME: the semantic layer (Semantic Types - STs), the analytical units used to transform between STs (Analysis Modules), the units of work flow (Analysis Chains), the work-flow processor (Analysis Engine), and the transportability layer (OME XML). An example of a complex work flow will also be presented illustrating how this system is used for automated image classification.