The Walker 256/B mammary carcinosarcoma implanted in male Fischer rats was used to evaluate bone resorption inhibitors. The model was improved by thyroparathyroidectomizing the animals to avoid counter-regulation by parathyroid hormone or calcitonin, by pair-feeding them, and by using 2 hour fasting calciuria as a resorption index to minimize the influence of differences in growth or in intestinal calcium absorption. Over a 10 day period, the control animals displayed a progressive increase of plasma calcium (Ca) and fasting urinary Ca excretion, a decrease of plasma phosphate, and an increase of urinary phosphate excretion. Osteocalcin did not change. The daily administration of dichloromethylene bisphosphonate (Cl2MBP) totally prevented the increase of fasting urinary Ca excretion, whereas plasma Ca remained at a higher level than thyroparathyroidectomized (TPTX) control rats. Osteocalcin decreased. Two new aminobisphosphonates, 4-amino-1-hydroxybutylidene-1,1-bisphosphonate (AHBuBP), and 6-amino-1-hydroxyhexylidene-1,1-bisphosphonate (AHHexBP) had a similar effect. The order of potency shown by the three bisphosphonates was similar to that reported using the metaphyseal bone density evaluation method in growing rats: AHBuBP greater than AHHexBP greater than Cl2MBP, the difference each time being one order of magnitude. The analysis of the relationship between urinary and plasma values in tumor-bearing animals suggested an increased renal tubular reabsorption of Ca and a decreased reabsorption of phosphate (Pi). Therefore, in this model of malignant osteolysis, urinary Ca excretion is the best marker for bone resorption.