Thanks for reply. I wrote a quick program and it seems to work.

>>>>While not a large table about 2000-3000 records it has losts of fields. He asked how hard
>>>>would it be do re-sequence the fields so they are sequnecial (i.e. nolonger as missing
>>>>values). Is there an efficent way of doing this? If so what is it?
>>>
>>>Would it work for you to make copies of those tables, associate them with a DBC, set cascading referential integrity on updates, make your changes, and then copy the data back to the original tables?
>>
>>I intended to work on a copy of the origninal table.
>>
>>My question is once I do that is the best way to resequence the key field which is defined
>>as a 9 character field, but is really a nine digit value with leading zeros, i.e. the simplest method
>>the comes mind is in a SCAN loop start at the first record and for each record replace the
>>current value of the key field with an integer value converted into a 9 character value with leading
>>zeros.
>>
>>Is this the best approach or is there a better one?
>>
>>The key field is not linked to any other table in any way (other fields are, but they are used
>>purely as decodes in reports).
>
>Oh, I assumed you meant that the key field appeared in other tables: that's why I suggested using referential integrity to make the updates. Yeah, your method above should work.