OBJECTIVE: To develop and validate a scoring system for selection of patients who should proceed to endocrinologic examinations of primary aldosteronism in newly diagnosed hypertensive patients. METHODS: A multivariate logistic regression analysis for primary aldosteronism was undertaken by use of seven possible primary aldosteronism markers, age less than 40 years, female sex, moderate-to-severe hypertension, hypokalemia, serum Na minus Cl at least 40 mmol/l, serum uric acid 237.92 μmol/l or less (4.0 mg/dl), and urine pH (U-pH) at least 7.0, in consecutive outpatients newly diagnosed with hypertension. The diagnostic criteria of primary aldosteronism were plasma aldosterone concentration-to-plasma renin activity ratio [ARR, (ng/dl)/(ng/ml per h)] at least 20 and at least one positive result in four types of challenge tests. RESULTS: Of 130 patients, 24 were diagnosed with primary aldosteronism. The area under the receiver operating characteristic curve (AUC) for a logistic model incorporating all possible primary aldosteronism markers was 0.73 [95% confidence interval (CI): 0.61-0.85]. Removing high U-pH, female sex, and hypokalemia from the full model decreased the AUC by 0.059, 0.035, and 0.011, respectively. We devised pH of urine, female sex, low serum K (PFK) score, in which one point each was assigned to high U-pH, female sex, and hypokalemia. The prevalences of primary aldosteronism in patients with 0, 1, 2, and 3 points were 11, 14, 42, and 60%, respectively. In external validation datasets (n = 106), AUC of PFK score was significantly higher than that of hypokalemia alone (0.73, 95% CI: 0.63-0.83 vs. 0.53, 95% CI: 0.44-0.63, P < 0.01). CONCLUSION: PFK score may be a better parameter than hypokalemia alone for identifying patients with a high probability of having primary aldosteronism.