Thanks, looks good to me, the performance was degrading quickly, this would probably fix it...

Thanks, that might solve my problem...

I didn't put the WHERE clause because it's very complex but it's working.

In fact, I had something like SELECT [IJK],[XC],[YC],[ZC] FROM [MSDP]; and it was taking 10 to 15 minutes to get the records. I tried to change it to SELECT TOP 200 [IJK],[XC],[YC],[ZC] FROM [MSDP] ORDER BY [IJK]. It's fast for the first 200 records. To get the following records, I use SELECT TOP 200 [IJK],[XC],[YC],[ZC] FROM [MSDP] WHERE [IJK] NOT IN (SELECT TOP 200 [IJK],[XC],[YC],[ZC] FROM [MSDP] ORDER BY [IJK]) ORDER BY [IJK]; the second 200 changes in a loop to 400,600 and so on. At the beginning, it's quite fast but the performances are degrading quite fast. Any help?

Hi everyone, I'm using a huge database of geostatistical blocks (million of records). I would like to read the records one by one and process only some. I tried the datareader but it's slowing terribly. Is there a read next or something like that in SQL? I'm programming in C#. Thanks in advance, Yvon