It seems commonly argued that evolution couldn't have happened in part because it violates the second law of thermodynamics. I really wish that I would stop seeing this. First, the second law has a technical meaning over and above 'systems tend to become more disordered with time'.
The second law of thermodynamics says that the entropy in systems increases with time. What this is is a statistical law about the availability of energy to do to work. Entropy is the change of heat/temperature change, and is defined in terms of the number of states a system can be in for a particular energy. The more states a system can be in with that energy the higher the entropy of the system. Examples include if you were to put a pot of boiling water in the center of the room. Over time the system, the room, will come to thermal equilibrium and the entropy has increased and is in fact maximized. What this means is there are no temperature differentials left which you could use to do mechanical work, say, and there are maximal arrangements that have that same energy. Compare that to the situation in which you had just put the pot of boiling water into the room, at the outset you could have potentially used the differences in temperature between the boiling water and surrounding air to drive some mechanical work. The concept of entropy extends to other systems that don't explicitly involve heat transfer also and may be easier to visualize. Suppose you have a swimming pool of white paint now drop in a bucket of red paint. At the very instant the bucket is dropped in entropy is minimalized, there are few ways of arranging the constituents of the pool to replicate that exact situation- red in one small spot, white everywhere else. Now wait for a while and come back, and the swimming pool is pink. Entropy is maximal, there are lots of ways of arranging the constituents of the swimming pool that will give you the same color.
There are a few things to note about this. The first is, as I stated earlier, this is a statistical rule. In the swimming pool you will have in fact localized regions where there are going to be a few more red paint molecules than white ones, but on the whole they are evenly mixed when the entropy of that system is maximized. It is logically possible that you will come back after seeing the pool pink to have the red re-congregate on its own in some corner, it's just... very very unlikely that would happen.
The second thing to note is that these systems are *left by themselves*. If you were to put a hot plate under the pot of boiling water in the first system (and say, open a window elsewhere) you could continue to push entropy down with time. Likewise with the pool, if you keep draining off paint at one end and feeding in white paint somewhere else, while adding red to a particular spot, you could keep that system from color equilibrium.
So the upshot of all this. The earth, as has been mentioned elsewhere, is not a closed or isolated system. Not only does the surface receive a constant inundation of energy from the sun (hot plate) the core creates heat which heats parts of the ocean, for instance. Such temperature differentials could, in principle, drive chemical reactions. The point is that the first life, for evolution to work, had to have some way of extracting energy out of the system (earth) in order to maintain its low entropy condition, and so long as you have that (i.e. hot plate, new paint) you could in principle according to the laws of thermodynamics have evolution occur.