Human atrial natriuretic peptide, known as carperitide, is approved for early relief of dyspnea in patients with acute heart failure (AHF). However, the diuretic effect of carperitide is sometimes insufficient for controlling volume overload. We investigated predictors for the carperitide response in patients with AHF. Forty-seven patients (age: 74 ± 10 years; left ventricular ejection fraction: 42.0% ± 15.9%) with AHF were enrolled and treated with carperitide monotherapy at a dose of 0.0125 μg/kg/min. Patients without sufficient diuresis (< 60 ml/h) or improvement of symptoms by 4 h after carperitide administration, despite increasing to twice the dose of carperitide and adding another agent, were defined as non-responders. Twenty-four (51%) patients were defined as responders and treated with low-dose carperitide monotherapy on the first day. Multiple logistic regression analysis showed that the response to carperitide monotherapy was independently predicted by serum creatinine levels and systolic blood pressure (SBP) on admission. The area under the receiver-operating characteristic curve for predicting the response to carperitide by SBP was 0.808 (95% confidence interval [0.686–0.930], sensitivity: 83.3%, specificity: 65.2%, cutoff value: 135 mmHg). Four (8.5%) patients developed asymptomatic transient hypotension. Worsening renal function occurred within 3 days of admission in three (6.4%) patients who received low-dose carperitide therapy. SBP and serum creatinine levels on admission might be useful for predicting the diuretic response to low-dose carperitide monotherapy in patients with AHF. Initial use of low-dose carperitide therapy does not have adverse effects on renal function.