OBJECTIVE To assess the relation of type 2 diabetes occurring earlier (age <55 years) versus later in life to the risk of cardiovascular death and to diabetes in offspring. RESEARCH DESIGN AND METHODS In the Framingham Heart Study, a community-based prospective cohort study, glycemic status was ascertained at serial examinations over six decades among 5,571 first- and second-generation participants with mortality data and 2,123 second-generation participants who initially did not have diabetes with data on parental diabetes status. We assessed cause of death in a case (cardiovascular death)–control (noncardiovascular death) design and incident diabetes in offspring in relation to parental early-onset diabetes. RESULTS Among the participants in two generations (N = 5,571), there were 1,822 cardiovascular deaths (including 961 coronary deaths). The odds of cardiovascular versus noncardiovascular death increased with decreasing age of diabetes onset (P < 0.001 trend). Compared with never developing diabetes, early-onset diabetes conferred a 1.81-fold odds (95% CI 1.10–2.97, P = 0.02) of cardiovascular death and 1.75-fold odds (0.96–3.21, P = 0.07) of coronary death, whereas later-onset diabetes was not associated with greater risk for either (P = 0.09 for cardiovascular death; P = 0.51 for coronary death). In second-generation participants, having a parent with early-onset diabetes increased diabetes risk by 3.24-fold (1.73–6.07), whereas having one or both parents with late-onset diabetes increased diabetes risk by 2.19-fold (1.50–3.19). CONCLUSIONS Our findings provide evidence for a diabetes subgroup with an early onset, a stronger association with cardiovascular death, and higher transgenerational transmission.