Statistical Control Group Experiment?

I want to set up a control group experiment of two groups the test and control, from a population of n= 11,000,000 individuals how big should my control groups be in other words how do I decide the sample size ?