Clusters of galaxies have long been an important cosmological tool. The next decade of observations promises to produce substantial new samples of galaxy clusters, detected by various means, and spanning a large range in redshift. Most cosmological diagnostics available from clusters involve the accurate measurement of cluster masses. Using N-body and hydrodynamic simulations, we revisit cluster mass estimation and consider issues important to obtaining masses from weak lensing, X-ray, and Sunyaev-Zel'dovich observations of clusters, and to using those masses to constrain the background cosmological model.