The natural history of stone disease in children is not well defined. We evaluated the clinical outcome in children with urinary calculi. An 8-year retrospective review of 129 pediatric patients with primary urinary lithiasis was performed. Age, renal versus ureteral stone location, stone size, spontaneous passage, recurrence and metabolic evaluation were considered. Patients were divided into groups 1-0 to 5, 2-6 to 10 and 3-11 to 18 years old. Of the 25 group 1 patients 17 (68%) had renal and 8 (32%) had ureteral stones. Of the 36 group 2 patients 13 (36%) had renal and 23 (64%) had ureteral stones. Of the 68 group 3 patients 12 (18%) had renal and 56 (82%) had ureteral stones. These differences in stone location according to age were not due to chance (p <0.0001). In groups 1 to 3 renal calculi an average of 6.7, 9.2 and 6.8 mm. spontaneously passed in 24%, 8% and 50% of cases, while ureteral calculi an average of 4.5, 3.5 and 3.2 mm. passed in 63%, 61% and 64%, respectively. The spontaneous passage rate of ureteral stones was consistent in the 3 age groups and for stone size up to 5 mm. Only 1 stone greater than 5 mm. passed spontaneously at any age. The incidence of identifiable metabolic abnormalities believed responsible for stone disease was 50% in groups 1 and 2, and 38% in group 3. In all age groups there was symptomatic and/or radiographic stone recurrence in a third of the patients with an identifiable metabolic abnormality, such as hypercalciuria, hypocitruria, renal tubular acidosis and so forth. In children 10 years or younger this incidence increased to 50%. Less than 10% of those with no identifiable metabolic disorder have had recurrent stones to date. Younger patients are more likely to present with renal calculi and less likely to pass these stones, probably due to the relatively larger stone burden and location. The passage rate for ureteral calculi is surprisingly consistent in all age groups with stones greater than 5 mm. rarely passing spontaneously. Half of the children 10 years or younger who present with urinary calculi have an identifiable metabolic disorder. Thus, all children with stones should undergo metabolic evaluation. In addition, these children are nearly 5-fold more likely to have recurrent stones than those with no identifiable metabolic disorder. Thus, they should be followed aggressively.