Primary aldosteronism (PA) is the most frequent form of secondary hypertension and results in an increased incidence of cardiovascular complications, independent of blood pressure. PA has two main causes: unilateral aldosterone-producing adenoma (APA, Conns syndrome) and bilateral adrenocortical hyperplasia (BAH). The distinction is important because adrenalectomy for APA has the potential for cure. In order to identify an APA adrenal vein sampling (AVS) is advocated by guidelines. AVS entails selective cannulation of the adrenal veins and sampling of blood of these for determination of aldosterone and cortisol. AVS is technically demanding and expensive and has therefore limited availability. A cheaper and simpler classic alternative to AVS is adrenal CT-scanning but the concordance between AVS and CT-scanning with regard to the presence of an APA is poor. The supposed superiority of AVS for selecting patients for adrenalectomy is mostly based on retrospective studies in which AVS results guided management and in which clinical follow-up was often lacking, all leading to a high risk of bias. We therefore performed a pragmatic randomised diagnostic trial, SPARTACUS (Subtyping PA: a Randomized Trial Comparing AVS and Computed Tomography Scan) in which we used clinical outcomes (antihypertensive medication use, blood pressure, biochemical cure, quality-of-life and costs) to determine the clinical value of AVS and CT in 200 patients with PA. The main findings are that clinical outcomes are similar regardless if AVS or CT is used to guide management. Both tests are imperfect in biochemical cure. Intriguingly the concordance between CT and AVS in the AVS arm was dismal. In my talk I will discuss the importance of the SPARTACUS design for diagnostic problems and the possible explanations for the findings on AVS performance and the potential for improvement. I will also discuss the implications for research and clinical management of PA.