Diabetes is characterized by excess levels of glucose and lipid in the blood, which have been proposed to directly and negatively impact islet β cells, exacerbating symptoms and promoting disease progression. Our understanding of “glucotoxicity” and “lipotoxicity” in human type 2 diabetes is limited, because most investigation has used either rodent islets or human islets in vitro. Based on these studies, numerous mechanisms have been proposed, but not tested, in human islets in vivo. To investigate the molecular mechanisms of excess glucose or lipid, we developed and employed three complementary models of nutrient excess, hyperglycemia (excess glucose), insulin resistance (excess lipid), or both, and examined direct effects on human islets in vivo. In response to hyperglycemia or insulin resistance, stimulated insulin secretion, the main indicator of β cell function, is severely reduced from human islet grafts. Surprisingly, in contrast to rodent islets, neither excess of lipid or glucose stimulated β cell proliferation or caused β cell apoptosis. In human islets in vivo, insulin resistance (1) blunted the unfolded protein response (UPR), (2) increased reactive oxygen species production and oxidative stress, (3) increased lipid droplets and amyloid deposits, and (4) reduced the key β cell transcription factors MAFB and NKX6.1. In contrast, hyperglycemia only blunted the UPR and reduced expression MAFB. Importantly, we found fundamental differences between human and mouse islets in response to identical metabolic conditions, such as stimulated insulin secretion, proliferative response, UPR induction, and transcription factor expression. To complement these studies, we performed a comprehensive, systematic, post-hoc analysis of variation of attributes and function across human islet preparations. Based on our data, we propose a model of glucotoxic and liptoxic consequences for human islets in vivo in which insulin resistance induces significantly more detrimental effects for human β cells in vivo than does hyperglycemia alone, but both conditions mechanistically converge to impair stimulated insulin secretion from human β cells in vivo. This work is the first to demonstrate and mechanistically investigate glucotoxic and lipotoxic consequences for human β cells in vivo, and the models developed will be useful to better understand and test interventions for gluco- and lipotoxicity.