Hi Dragan,

That "significant improvement" doesn't exist. I converted a table with 50,000+ records, then ran the same query against it and the original. The query took the 50,000+ and SUM()med a number of fields into 3 records. I ran it a dozen times, stored the elapsed times, and threw out the high and low results. The final outcome: With converted fields, average elapsed time 1.89 seconds. Without converted fields, average elapsed time 1.22 seconds. Three (acutally less) whole hundredths of a difference.