I'm running a forecasting system across a number of access databases. Unfortunately one of the databases has undergone heavy corruption of the data, but not in any logical method. I have run queries to find what extent data differs in the matrices and found some seventy lines have become illogically corrupted. The database itself still operates. I have checked for migration across parts (lines) but there is none. I have checked for trimming of numbers (i.e. 200 becoming 2, or 205 becoming 5). It's almost as if someone has lifted the file and replaced everything with a random number table. I am now at a loss as to how this has happened. Has anyone any ideas as to what could have caused this. The data matrices is of some 80,000 items which are manipulated by thirty different access databases. I fear it may just be human error, or just the size and structure of the project has reach disproportional levels and realy should have been ported to SQL server many months ago. But they don't listen to me do they!