How does having unequal sample sizes effect my statistical conclusions?

I have 2 groups, one who received treatment and a control group. About 12 variables were measured in each group. The Treatment group has about 65 subjects, the control group only has 25-30. Is it better (statisticaly speaking) to have the groups as is (65:30) or should I take a random sub-sample of the Treatment group and make it even (30:30)? Seems to me like more=better, but I'm not sure if that is true or not.....