Like all scientific theories, the theory of evolution must make predictions that are testable. Testing a prediction means that you predict what evidence you should find if the theory is correct, and then you look to see if you actually find that evidence. So what does the theory predict we should find, and do we actually find it?
First of all, the theory predicts we should see examples of populations of one species evolving into new species. Do we? Yes we do. http://www.talkorigins.org/faqs/faq-spec…
http://en.wikipedia.org/wiki/Ring_specie…
Secondly, the theory predicts the existence of a Tree of Life. The Tree of Life is a family tree, of which every single living thing that has ever existed on the planet is a member. The existence of the Tree of Life (if true) would mean that the differences between living things would be organized into a nested hierarchy, such that two recently separated species would share numerous characteristics, of which a portion would also be shared by all species sharing a more distant common ancestor, of which a portion would also be shared by all species sharing an even more distant common ancestor; and so on. The Tree’s existence would also mean that two species branching from a common ancestor should be located in places that populations of the ancestor could have reached. It also means that the fossils we find should fit into the general pattern of hierarchical similarity, location, and timeframe established by the Tree. Finally, the hierarchical patterns of similarity established by present and past species on the Tree should be roughly the same regardless of whether we are comparing morphological characteristics, or genes, or non-coding DNA, or endogenous retroviruses, or proteins. We say “roughly the same” rather than “exactly the same” because various processes such as convergent evolution or fluctuating population sizes can somewhat throw off the hierarchical patterns established by the different traits being compared.