The objective of this study was to compare 18F-PSMA-1007 PET/CT and 18F-fluorocholine PET/CT for the localization of prostate cancer (PCa) biochemical recurrence. Methods: This prospective, open-label, randomized, crossover multicenter study included PCa patients with prior definitive therapy and suspected PCa recurrence. All men underwent both 18F-PSMA-1007 PET/CT and 18F-fluorocholine PET/CT (102 received 18F-PSMA-1007 PET/CT first and 88 received 18F-fluorocholine PET/CT first). All images were assessed independently by 3 readers masked to all clinical information using a 3-point qualitative scale (0 = no recurrence, 1 = undetermined, and 2 = recurrence). Patients were monitored for approximately 6 mo. An independent panel with a urologist, radiologist, and nuclear physician reviewed all clinical data, including imaging and response to therapy, but were masked regarding PET/CT information; acting in consensus, they determined a patient-based and region-based composite standard of truth for PCa lesions. The "correct detection rates" for PCa lesions on a patient basis for each radiopharmaceutical were compared for the 3 readers individually and for the "average reader." Secondary objectives included determining whether PET/CT findings affected diagnostic thinking (impact of a test result on posttest vs. pretest probability of a correct diagnosis), therapeutic decision making (description and quantification of impact of diagnostic information gained with both radiopharmaceuticals on patient management), and adequacy of management changes. Results: A total of 190 patients were included. The primary endpoint was met. The overall correct detection rates were 0.82 for 18F-PSMA-1007 and 0.65 for 18F-fluorocholine (P < 0.0001) when undetermined findings were considered positive for malignancy and 0.77 and 0.57, respectively (P < 0.0001), when undetermined findings were considered negative for malignancy. A change in diagnostic thinking due to PET/CT was reported in 149 patients; 18F-PSMA-1007 contributed more than 18F-fluorocholine in 93 of these patients. In 122 patients, PET/CT led to an adequate diagnosis that benefited the patient; 18F-PSMA-1007 contributed more than 18F-fluorocholine in 88 of these patients. Conclusion: 18F-PSMA-1007 PET/CT is superior to 18F-fluorocholine PET/CT for the localization of PCa recurrence. Decision making was more beneficial when based on 18F-PSMA-1007 PET/CT results.