We examined the effectiveness of chlorthalidone or magnesium hydroxide in the prevention of recurrent calcium oxalate kidney calculi. In a double-blind random allocation design daily dosages of 25 or 50 mg. chlorthalidone, 650 or 1,300 mg. magnesium hydroxide, or an identical placebo were administered. All groups showed significantly decreased calculous events compared to the pretreatment rates. During the trial 56.1 per cent fewer calculi than predicted developed in the placebo group (p less than 0.01), whereas the groups receiving low and high dosage magnesium hydroxide showed 73.9 and 62.3 per cent fewer calculi, respectively (p less than 0.001 and less than 0.01, respectively). Chlorthalidone treatment resulted in a 90.1 per cent decrease from predicted rates and both dosages yielded similar results. When the treatments were compared chlorthalidone was significantly better than the placebo or magnesium hydroxide (p less than 0.01). The large decreases in calculous events seen when placebo or ineffective therapy was given underscore the positive treatment bias that occurs when historical controls are used and they demonstrate the need for proper experimental design.