It is important to test only one variable when conducting an experiment because that one variable will only have one change. You want to know whether or not that one change is what caused the outcome of the experiment. If you use too many you don't know what really caused the outcome.