I have a proram that extracts images from an Oracle database. I query on some data attributes and a blob column. If there are more than a 1000 images to extract an issues araises. The pattern is pretty constaint at 1000 good extracted images then the next 400 image written to the filesystem are a duplicate of the last good image to be extracted. Then next 100 images are fine again. The next 400 are dupes of the last good image form the 100. This pattern repeats. I have seen only one case were the pettern changed and the 400 dupelicate image were only 200. This issue happens on any computer we run from. I have 2gigs memory and a 1.5gig swap file. I feel this must be something to do with the ADO recordset or the Oracle connection. Something external of my code but can't figure it out. Could someonr take a look and give me some direction?