I have a bunch of XML files...of unknown structure/origin - but all having the same schema. I need to write a tool to transform this XML document (en-masse) into a database format. The XML documents all have the same structure, but the tool needs to be flexible enough to work with any XML document/schema. I've got the tool opening the XML file into a dataset but here is where I come unstuck. I've got the user selecting the MDB to import the file content to and want my application to analyse the database to see whether the datatables from the dataset exist in the database, if they do, then append the data for the corresponding table, else create the table dynamically from the schema structure of the datatable within the dataset. Given that the dataset is dynamically built at the time the XML document is loaded, how would I go about dynamically creating the necessary tables in the database and/or appending the data from each datatable within the dataset?