Hmmmm. I'm thinking the explicit select on primary key would be faster. The Find method of the ADO recordset only searches in one direction - Forward or Backward. When you find a particular record using Find, the record pointer is set to that record. Your next Find will search from that point. For what you are trying to do that obviously doesn't help since the value you want may be "behind" the record pointer.

The way you make the Find do what you want it to do is to provide the SearchDirection and Start parameters like this:

oRS.Find "pkey = 1111", SearchDirection:=adSearchForward, Start:=adBookmarkFirst

Now this would effectively be the same as:

oRS.MoveFirst
oRS.Find "pkey = 1111"

Two ADO operations for the price of one! What a deal! Well, not so much. Basically you're just doing a SCAN. If your ADO connection stays open thoughout the loop's iterations I think that doing an explicit query against the primary key would be faster. Of course on such a small recordset the difference may be negligible.

>Hey folks,
>
>General ADO question here: Is there a significant speed difference between using the Recordset.Find method on a recordset with about a dozen rows with a single key field created using .Open("SELECT * from table") and .Open("SELECT * FROM table WHERE pkey = 1111")?
>
>I ask this in the context of a lookup table that needs to be hit up about 100 times in a big loop. Right now, everything is "Open" where I SELECT for the single key and I'm thinking of changing that if some benefit will be derived by opening the entire table all at once and using Find in the loop.
>
>It's a big programming task and I was looking for some informed opinions first.
>
>Thanks!