Study design: A matched case-control study. Objectives: To clarify the influence of urine specific gravity and drinking water quality on the formation of urinary stones in persons with spinal cord injury (SCI). Setting: A rehabilitation center within a university hospital. Methods: Between 1992 and 1998, 63 stone cases (31 kidney, 27 bladder, and five both) and 289 age-duration-matched controls were recruited from a cohort of SCI patients enrolled in an on-going longitudinal study. Data on urine specific gravity and other characteristics of study participants were retrieved from the database and medical charts. Community water supply information was provided by the Alabama Department of Environmental Management. Multivariable conditional logistic regression analysis was performed to evaluate the association with stone formation. Results: SCI individuals who had urinary stones were more likely than control subjects to use indwelling catheters and have decreased renal function. The occurrence of stones was not significantly related to gender, race, severity of injury, urinary tract infection, nor urine pH. After controlling for the potential confounding from other factors, a continuously increasing stone occurrence with increasing specific gravity was observed (P= 0.05); this association was stronger for kidney (odds Ratio [OR] = 1.8 per 0.010 g/cm3) versus bladder stones (OR = 1.2) and for recurrent (OR = 2.0) versus first stones (OR = 1.5). Increased water hardness was not significantly associated with a decreased stone occurrence. Conclusions: Study results suggest that maintaining urine specific gravity below a certain level might reduce the occurrence of urinary stones. This could be easily achieved by using a dipstick for self-feedback along with appropriate fluid intake. For persons with SCI who are at an increased risk of a devastating stone disease, this prophylactic approach could be very cost-effective; however, this requires further confirmation.