The Find methods are also reasonably efficient with snapshot-type Recordset objects that have a small number of records. In this case, the find criteria are resolved by searching the data that has already been downloaded into the snapshot on the local machine. Of course, if the snapshot is not yet fully populated, the time required to retrieve the data increases the search time. This means that Find operations that fail will be slow if all the records need to be retrieved (unlike the previous dynaset case in which no further retrieval occurs if the Find method fails).