Modeling studies are an important way for climate scientists to explore the dynamics of past climate states. However, nearly all climate models were initially developed to examine modern climate conditions. How effective are these models at accurately representing past climate states? Collaboration between climate researchers who generate paleoclimate data and those who employ paleoclimate models can help answer this question. Proxy records provide data that paleoclimate modelers can use to validate their models. In contrast, modeling results can help data generating researchers test the plausibility of their ideas about past climate dynamics. I have a collaboration with Dr. Alan Haywood of the University of Leeds (UK), whose work has provided a good test of some of my ideas about climate conditions in the North Atlantic Ocean during the Pliocene. I also have done work with paleoclimate modelers Dr. Alexey Fedorov and Dr. Chris Brierley from Yale University. Work integrating paleoclimate modeling result from with data generated by myself and others was published in Science magazine in Brierley et al. 2009. We are currently working on synthesis paper for the Pliocene that incorporates both data and modeling results.