Forty adult male Wistar rats were placed in metabolic cages on a Ca-deficient diet (0.1%) for 7 days and then on a Ca-deficient, Na-oxalate (NaOx) enriched diet (20 mg/100 g) for another 14 days. The animals were subdivided into three groups receiving three different types of mineral water: group I (n = 13), Badoit (Ca 222 mg/l); group II (n = 14), Contrexéville (Ca 467 mg/l); and group III (n = 13), Evian (Ca 78 mg/l). Another series of 25 rats (group I, n = 9; group II, n = 8; group III, n = 8) underwent the same study protocol, except that they received a normal Ca diet (1%). On the low-Ca diet, urinary Ca-Ox monohydrate (COM) crystals were observed only under the Na-Ox diet, with a mean crystal number significantly greater in group III (16.7 +/- 4.5 crystals/mm3) than in group I or II rats (2.5 +/- 1.5 or 4.1 +/- 1.5 crystals/mm3, respectively). Urinary Ca concentrations decreased in all groups (P < 0.001) under the Na-Ox diet, while urinary oxalate concentrations increased in all groups (P < 0.001). On the normal Ca diet, COM crystal excretion was observed only with the Na-Ox-enriched diet, but in this case feeding the Na-Ox diet did not modify urinary oxalate excretion. Ca/Ox ratio was significantly lower under 0.1% Ca diet than under normal Ca diet, related with the type and the number of crystals observed, demonstrating that assessment of crystalluria can provide an index of disease severity. Moreover, the hardness of the drinking water influences urinary COM crystal excretion only under a low-Ca, oxalate-rich diet, suggesting that the total calcium intake rather than the water calcium content is an important factor in the occurrence of Ca-Ox nephrolithiasis.