Last time I worked on a CLR function for SQL Server, I was unable to find a way to handle an exception and pass an error back through the pipe to SQL Server. Now this was a scalar function versus a table value function, but I'm not sure you can handle the error a pass an exception back through the pipe to SQL Server on in a CLR function. You may be stuck with getting the .NET exception back in TSQL. You can do handle the exception and pass an error back through the pipe in a CLR Stored Procedure, however.