>Sorry, it should have been:
>
>x = asc(fread(mfile,1)) + (asc(fread(mfile,1)) * 256) + ;
>>(asc(fread(mfile,1)) + (asc(fread(mfile,1)) * 256)) * 256 * 256
>
Paul,
I know for a fact that it is floating point. I have the full doc on the files. In a trip to the library last night I found that floating point numbers contain an exponent plus the number. Something like scientific notation but not quite. The books I found said the method is system dependent. The data I am reading was written with a Borland C++ program (I think). The four bytes of the number are different from four bytes of an intiger in that with an intiger the left bytes have values such as 0A 12 00 00 whereas the floating point values are on the right side 00 00 CC 41. I am assuming that the exponent is in x number of bits in the rightmost byte.

John.