Open the table and its companion files
Keep a copy of the original DBF before starting. If the table uses memo fields, retain the matching DBT or FPT with the same base filename. The DBF usually stores a pointer to the memo, not the memo’s text. A table that opens without its sidecar can therefore appear to have empty notes even though its ordinary columns are intact.
Choose the supported files and an active/deleted record selection, then run the conversion. The generated example contains an identifier 00001234, the numeric text 9007199254740993.2500, a real DBT note and one deletion-marked record. It demonstrates the distinctions that a successful migration should retain rather than treating every visible value as an ordinary spreadsheet number.
Use the workbook’s explicit types
The Data worksheet stores DBF numeric values as text. Excel normally keeps about fifteen significant decimal digits in a numeric cell, so importing a long identifier or precise stored amount as a number can change its value. Text cells retain leading zeros and the original decimal spelling. Logical fields become booleans; dates and date-times become ISO text without an invented timezone.
The Schema worksheet records each source field’s DBF type, width, decimal count and nullable flag. Source record numbers and deletion flags accompany every exported row. Those references let you compare an unexpected value with its location in the original table. The workbook does not recreate indexes, application formulas, relationships or business rules.
Review the exception paths
Read the Issues worksheet before using the data. An unsupported binary memo, malformed value or missing memo pointer produces an unresolved cell with a reason and its stored bytes. It does not become an invented empty note. JSON keeps explicit nulls and the complete issue record, while a spreadsheet displays such cells as blank.
Excel permits at most 32,767 characters in one cell. Longer decoded text is split into ordered chunks on the Long text worksheet, and the Data cell points to those chunks. CSV and JSON retain the full value. Preserve the originals and inspect representative records in the destination application before treating the migration as complete.