With rising incidence and prevalence of type 2 diabetes, prevention including identification of prospective biomarkers becomes increasingly relevant. Although ketone bodies recently received a renewed interest as potential biomarkers, data linking these metabolites to diabetes risk are scarce. Therefore, the present prospective study investigated a potential association between fasting ketone bodies and incident type 2 diabetes in the general population. This study from the PREVEND cohort included 3,307 participants from the general population initially free of diabetes or impaired fasting glucose. Baseline fasting ketone body concentrations were measured by nuclear magnetic resonance spectroscopy. One hundred twenty-six participants (3.8%) developed type 2 diabetes during a median (interquartile range) follow-up of 7.3 (6.3-7.6) years. In Kaplan-Meier analysis, sex-stratified ketone body levels strongly positively associated with incident type 2 diabetes, which was confirmed in Cox regression analyses adjusted for several potential confounders. There was no significant interaction by sex. Both 3-β-hydroxybutyrate and acetoacetate+acetone individually associated with incident type 2 diabetes. In conclusion, fasting plasma ketone body levels are strongly positively associated with incident type 2 diabetes in the general population independent of several other recognized risk factors. These results may have important implications for diabetes prevention including dietary strategies. The identification of biomarkers that predict type 2 diabetes is increasingly relevant for personalized medicine strategies. Data regarding ketone bodies and incident type 2 diabetes are scarce. This study shows that ketone bodies, either combined or as individual subspecies, are strongly associated with incident type 2 diabetes in the general population, independent of potential confounders. These results may have important implications for diabetes prevention including dietary strategies.