It is unlikely that the user connections are causing the differences in size. It is more likely from the replication
settings and what is actually replicated. If you are not replicating design elements, ACLs, db headers and
view indexes (which don’t replicate anyway) then the db’s will be a different size.
Another way to see that the DBs are the same size is to look at the document count on the two databases.
They should be within a few documents of each other. To see this (if the db is indexed) go to the properties
of the db and click on the 6th tab over from the left. Click on count unindexed documents. Do this on both
databases and compare thecounts. They should be very close.
The most likely culprit here (mentioned above) is view indicies. An index takes up space, and in a very large DB, it can get quite large. An index is built when that view is accessed. If you replicate a DB to another server where people do not use it, those indicies are not built. The doc counts should be quite similar, but size can vary greatly due to indicies.