It eliminates all forms of bias.

It reduces variability by controlling for known nuisance variables.

It allows for a larger sample size.

It removes the need for a control group.