Database replication is the act of copying, or replicating,
data from one table or database to another table or database.
Using this technology,
you can distribute copies of an entire database to multiple systems throughout your company,
or you can distribute selected pieces of the database.
When SQL Server replication technology is used,
the task of copying and distributing data is automated.
No user intervention is needed to replicate data once replication has been set up and configured.
Because the data replication and processing is done from within a SQL Server database,
there is additional stability and recoverability.
If a failure occurs during replication (or while any other SQL Server transaction is being performed),
operations resume at the point of failure once the problem is fixed.
Because of this, many people prefer replication to other methods of moving data between systems.
You have many options for configuring replication on your network.
For example, you can specify how much or how little data will be replicated.
You can specify whether the replicated copies will be read-only or can be modified.
And you can specify how often the data is to be replicated.
Source: Microsoft SQL Server 2000 Administrator's Companion Book