The Heat Shock Protein 90 (Hsp90) is an essential and highly conserved chaperone that facilitates the maturation of a wide array of client proteins, including many kinases. These clients in turn regulate a wide array of cellular processes, such as signal transduction, and transcriptional reprogramming. As a result, the activity of Hsp90 has the potential to influence physiology, which in turn may influence the ability to adapt to new environments. Previous studies using a deep mutational scanning approach, (EMPIRIC) identified multiple substitutions within a 9 amino acid substrate-binding loop of yeast Hsp90 that provides a growth advantage for yeast under elevated salinity conditions and costs of adaptation under alternate environments. These results demonstrate that genetic alterations to a small region of Hsp90 can contribute to evolutionary change and promote adaptation to specific environments. However, because Hsp90 is a large, highly dynamic and multi-functional protein the adaptive potential and evolutionary constraints of Hsp90 across diverse environments requires further investigation. In this dissertation I used a modified version of EMPIRIC to examine the impact of environmental stress on the adaptive potential, costs and evolutionary constraints for a 118 amino acid functional region of the middle domain of yeast Hsp90 under endogenous expression levels and the entire Hsp90 protein sequence under low expression levels. Endogenous Hsp90 expression levels were used to observe how environment may affect Hsp90 mutant fitness effects in nature, while low expression levels were used as a sensitive readout of Hsp90 function and fitness. In general, I found that mutations within the middle domain of Hsp90 have similar fitness effects across many environments, whereas, under low Hsp90 expression I found that the fitness effects of Hsp90 mutants differed between environments. Under individual conditions multiple variants provided a growth advantage, however these variants exhibited growth defects in other environments, indicating costs of adaptation. When comparing experimental results to 261 extant eukaryotic sequences I find that natural variants of Hsp90 support growth in all environments. I identified protein regions that are enriched in beneficial, deleterious and costly mutations that coincides with residues involved in co-chaperone-client-binding interactions, stabilization of Hsp90 client-binding interfaces, stabilization of Hsp90 interdomains and ATPase chaperone activity. In summary, this thesis uncovers the adaptive potential, costs of adaptation and evolutionary constraints of Hsp90 mutations across several environments. These results complement and extend known structural and functional information, highlighting potential adaptive mechanisms. Furthermore, this work elucidates the impact environment can have on shaping Hsp90 evolution and suggests that fluctuating environments may have played a role in the long-term evolution of Hsp90.