Reading the previously mentioned page closely I found section that contains conversion functions. http://www.merlyn.demon.co.uk/programs/floatval.pas The procedure EightToSix is probably what you are looking for Jan