After long-term therapy, some patients with systolic heart failure (HF) display improved left ventricular (LV) function over time, a response that is associated with improved long-term outcomes. To investigate predictors of improved LV function in an ethnically diverse HF cohort, we selected 71 patients with HF who had baseline ejection fractions (EF) <40%, follow-up EFs > or =50%, and >20% increases on follow-up echocardiography performed > or =6 months later. Their clinical features were compared with 142 age- and gender-matched control patients with baseline EFs <40% and no change or worse EFs on follow-up echocardiography. The baseline EFs were similar between patients and controls. Compared with controls, patients had a lower prevalence of diabetes mellitus (19.7% vs 36.6%, p = 0.01), a lower prevalence of an ischemic cause of disease (8.4% vs 35.2%, p <0.001), but a higher prevalence of a hypertensive cause of cardiomyopathy (29.6% vs 12%, p <0.001). Fewer patients than controls used aspirin (p = 0.04) or statins (p = 0.001) or had previous cardiac procedures (p = 0.009). In a multivariate conditional logistic regression model adjusting for age, gender, disease cause, statin use, cardiac procedures, change in heart rate, and follow-up time, hypertensive etiology was most strongly associated with improved LV function (adjusted odds ratio 9.73, 95% confidence interval 1.44 to 52.76, p = 0.02). In conclusion, patients with hypertensive causes of HF are more likely to demonstrate improved LV function over time than patients with ischemic causes of HF. Because long-term prognosis and indication for defibrillator implantation may be affected by changes in LV function, the present study provides additional support for the importance of evaluating the cause of HF to guide management.