Current practice recommends metabolic evaluation of patients who have formed multiple renal stones, but not those with one stone or temporally remote stones. This presumes that recentness and recurrence imply greater risk of new future stones. We hypothesize that number of stones reflects how long patients are permitted to form stones untreated, and that forming more stones, itself, raises risk of future stones despite treatment. Our report is a retrospective analysis of 371 male patients selected from a comprehensive clinical and laboratory data base containing 2,527 patients with nephrolithiasis. Before treatment, number of stone events rises with time of observation, and rate of stone event occurrence is constant or falls. During treatment, relapse is correlated with number of pretreatment stones. Life table analysis showed increasing relapse for patients grouped into those with one, two, and three or more stones. Even though number of stones seems controlled by the interval of observation before treatment, more stones predict higher relapse during treatment. Perhaps by leaving nuclei of crystals as residues, stones appear to promote new stones, and the practice of waiting while patients declare themselves multiple stone formers may not always be the best.