Pulse pressure (PP) remains an elusive cardiovascular risk factor with inconsistent findings. We clarified the prognostic value in patients with type 2 diabetes, chronic kidney disease (CKD) and anemia in the Trial to Reduce cardiovascular Events with Aranesp (darbepoetin alfa) Therapy. In 4038 type 2 diabetes patients, darbepoetin alfa treatment did not affect the primary outcome. Risk related to PP at randomization was evaluated in a multivariable model including age, gender, kidney function, cardiovascular disease (CVD) and other conventional risk factors. End points were myocardial infarction (MI), stroke, end stage renal disease (ESRD) and the composite of cardiovascular death, MI or hospitalization for myocardial ischemia, heart failure or stroke (CVD composite). Median (interquartile range) age, gender, eGFR and PP was 68 (60-75) years, 57.3% women, 33 (27-42) ml min(-1) per 1.73 m2 and 60 (50-74) mm Hg. During 29.1 months (median) follow-up, the number of events for composite CVD, MI, stroke and ESRD was 1010, 253, 154 and 668. In unadjusted analyses, higher quartiles of PP were associated with higher rates per 100 years of follow-up of all end points (P⩽0.04), except stroke (P=0.52). Adjusted hazard ratios (95% confidence interval) per one quartile increase in PP were 1.06 (0.99-1.26) for MI, 0.96 (0.83-1.11) for stroke, 1.01 (0.94-1.09) for ESRD and 1.01 (0.96-1.07) for CVD composite. Results were similar in continuous analyses of PP (per 10 mm Hg). In patients with type 2 diabetes, CKD and anemia, PP did not independently predict cardiovascular events or ESRD. This may reflect confounding by aggressive antihypertensive treatment, or PP may be too rough a risk marker in these high-risk patients.