Description: Scientific research frequently leads to complex mathematical models. To explore such models, researchers often turn to Monte Carlo methods, which are mathematical experiments that use random numbers. Monte Carlo experiments have become useful tools in physics, finance, and statistics, as well as in other branches of science. To run these experiments, one typically wants a computer to generate millions (or billions) of random numbers --- a task which seems to contradict a basic characteristic of computers. I will give some examples of Monte Carlo in action, and also describe how we get the random numbers that we need.