Introduction: Resistant hypertension is inadequately controlled blood pressure (BP) despite treatment with at least three BP-lowering drugs. A popular hypothesis is that resistant hypertension is due to excessive Na+-retention, and that 'further diuretic therapy' will be superior to alternative add-on drugs. Methods and analysis: Placebo-controlled, random crossover study of fourth-line treatment when added to standard (A+C+D) triple drug therapy: ACE inhibitor or Angiotensin receptor blocker (A) +Calcium channel blocker (C)+Diuretic (D). Patients (aged 18-79 years) with clinical systolic BP≥140 mm Hg (135 mm Hg in diabetics) and Home BP Monitoring (HBPM) systolic BP average ≥130 mm Hg on treatment for at least 3 months with maximum tolerated doses of A+C+D are randomised to four consecutive randomly allocated 12-week treatment cycles with an α-blocker, β-blocker, spironolactone and placebo. The hierarchical coprimary end point is the difference in HBPM average systolic BP between (in order) spironolactone and placebo, spironolactone and the average of the other two active drugs, spironolactone and each of the other two drugs. A key secondary outcome is to determine whether plasma renin predicts the BP response to the different drugs. A sample size of 346 (allowing 15% dropouts) will confer 90% power to detect a 3 mm Hg HBPM average systolic BP difference between any two drugs. The study can also detect a 6 mm Hg difference in HBPM average systolic BP between each patient's best and second-best drug predicted by tertile of plasma renin. Ethics and dissemination: The study was initiated in May 2009 and results are expected in 2015. These will provide RCT evidence to support future guideline recommendations for optimal drug treatment of resistant hypertension.