The article investigates the relationship between protein connectivity and evolutionary rates in the yeast *Saccharomyces cerevisiae*. The authors compiled a list of 3541 protein interactions involving 2445 different yeast proteins and estimated evolutionary rates by comparing orthologous sequences between yeast and the nematode *Caenorhabditis elegans*. They found a significant negative correlation between the number of protein interactors and evolutionary rate, suggesting that proteins with more interactors evolve more slowly. This correlation is not due to the proteins being more important to the organism but rather because a greater proportion of the protein is directly involved in its function. The authors also tested alternative hypotheses, including the effect of protein fitness on evolutionary rates, and concluded that coevolution, where substitutions in one protein lead to reciprocal changes in interacting partners, is the primary explanation for the observed similarity in evolutionary rates among interacting proteins. The findings have implications for understanding the constraints on protein evolution and the reliability of large-scale protein interaction datasets.The article investigates the relationship between protein connectivity and evolutionary rates in the yeast *Saccharomyces cerevisiae*. The authors compiled a list of 3541 protein interactions involving 2445 different yeast proteins and estimated evolutionary rates by comparing orthologous sequences between yeast and the nematode *Caenorhabditis elegans*. They found a significant negative correlation between the number of protein interactors and evolutionary rate, suggesting that proteins with more interactors evolve more slowly. This correlation is not due to the proteins being more important to the organism but rather because a greater proportion of the protein is directly involved in its function. The authors also tested alternative hypotheses, including the effect of protein fitness on evolutionary rates, and concluded that coevolution, where substitutions in one protein lead to reciprocal changes in interacting partners, is the primary explanation for the observed similarity in evolutionary rates among interacting proteins. The findings have implications for understanding the constraints on protein evolution and the reliability of large-scale protein interaction datasets.