Computers are becoming increasingly fast, making it possible to perform simulations of macromolecules on timescales that were previously inaccessible. Questions have arisen concerning how well we are keeping up with computer power and the state of the art with respect to long molecular dynamics simulations in solvent. More importantly, however, simulations of macromolecules are performed to aid the understanding of biochemical phenomena. So, what are we learning from longer simulations and are they providing reliable insight into protein dynamics, conformational behavior and function?