Purpose It is challenging to differentiate unilateral aldosterone-producing adenoma (APA) from bilateral idiopathic adrenal hyperplasia (IAH) and nonfunctional adrenal adenoma (NFA) in primary aldosteronism (PA). In a first primarily ex vivo study detection, CXC chemokine receptor type 4 (CXCR4) expression has been shown to be a valuable tool for the detection of APA. In this study, we aimed to clinically evaluate CXCR4 imaging with 68 Ga-pentixafor PET/CT for detecting APA. Methods We prospectively recruited 36 patients with clinical suspicion of PA. All patients underwent 68 Ga-pentixafor PET/CT. Positive lesions were defined based on higher tracer uptake in adrenal nodular(s) shown on CT than the normal adrenal. These lesions were referred for adrenalectomy subsequently. All patients received clinical follow-up. Semi-quantitative analysis using maximum standardized uptake value (SUV max ), lesion-to-liver ratio (LLR), and lesion-to-contralateral ratio (LCR) has also been performed. PET/CT results were correlated with clinical presentation and follow-up. Results Thirty-nine adrenal lesions in 36 patients were found; 25 APA, 4 IAH, and 10 NFA according to histopathology and clinical assessment. Sensitivity, specificity, and accuracy of 68 Ga-pentixafor PET/CT in distinguishing APA by visualization were 100%, 78.6%, and 92.3% respectively. The SUV max of APA (21.34 ± 9.41, n  = 25) was significantly higher than that of non-APA lesions (6.29 ± 2.10, n  = 14, P  < 0.0001). An optimal threshold of SUV max  = 11.18 was determined for predicting APA with a sensitivity of 88.0%, specificity of 100%, and an accuracy of 92.3%. A cutoff value for LCR of 2.12 yielded a sensitivity of 100% and a specificity of 92.9%, whereas a cutoff value for LLR of 2.36 reached at both 100% of sensitivity and specificity. All patients with (removed) positive lesions benefited from surgery. Conclusion 68 Ga-Pentixafor PET/CT may be used to non-invasively detect APA in PA patients.