DonQuichote is right, not much you can do for text files except what he suggests.
This isn't
strictly true, because you could open the text file as a BINARY file and overwrite any bytes you choose. The problem is you can only write back exactly the number of bytes that were there originally. You can't make a new "first line of text" that is longer or shorter than the original. Then there is the problem of trying to locate something like the 256th record of the file. There is no direct way to do this, you'd need to scan the file from the beginning, counting newline markers (vbCrLf) as you go. Yuck!
Of course fixed-length record RANDOM files are another story altogether. They solve all of your problems except when you need to delete a record from the original file. Plus, you need a properly formatted file in the 1st place, and text files almost always are
not in the form you need for RANDOM access.
I'm not sure
what that commercial for the Scripting.FileSystemObject ("FSO"

class was all about. VB6 has the same I/O statements that VB5 did. Yes, there are some issues with the older-style VB I/O. Yes, Microsoft
does recommend that people use the FSO instead of native VB I/O for many applications. But the FSO has many limitations that VB I/O can overcome: for example FSO doesn't support COM: port I/O, or RANDOM file I/O, or BINARY I/O. It only handles filesystem stream files.
I use FSO a lot myself, but I use VB I/O almost as often. It can do so much more! And the FSO isn't going to help you edit the first record of a text file either, unless you do something like DonQuichote suggested in the first place.
So much depends on what you are really trying to do.