Gudrun, I can't say with certainty that OCR processing a file that's already got a text layer will cause the problems you're seeing, but it has got to be "a bad thing" almost on the face of it. OCR is, by comparison, a very poor substitute for materials actually created with a text layer in them, and every downloadable tax form or instruction set I've ever dealt with is searchable already because it was created that way. What happens when you attempt to overlay that, well, I don't know.
All I can recommend is downloading again and doing all filling in with Adobe Reader DC. I know that one can use PDF XChange Viewer, but I've seen enough now with JAWS not playing nicely (or at all) with it at times that I'd stick with Adobe Reader. These forms should be not only fillable, but able to be saved with the information you've filled in using Adobe Reader DC. Brian
