The aim of this study was to compare the efficacy of trisacryl gelatin microspheres versus polyvinyl alcohol particles (PVA) in the preoperative embolization of bone neoplasms, on the basis of intraoperative blood loss quantified by the differences in preoperative and postoperative hematic levels of hemoglobin, hematocrit and erythrocytes count. From January 1997 to December 2002, preoperative embolization of bone tumors (either primary or secondary) was carried out in 49 patients (age range 12/78), 20 of whom were treated with trysacril gelatin microspheres (group A) and 29 with PVA particles (group B). The delay between embolization and surgery ranged from 1 to 13 days in group A and 1 to 4 days in group B. As used in international protocols, we considered hematic levels of hemoglobin, hematocrit and erythrocytes count for the measurement of intraoperative blood loss then the differences in pre- and postoperative levels were used as statistical comparative parameters. We compared the values of patients treated with embospheres (n = 10) and PVA (n = 18) alone, and patients treated with (group A = 10; group B = 11) versus patients treated without other additional embolic materials in each group (group A = 10; group B = 18). According to the Student's t-test ( p < 0.05), the difference of hematic parameters between patients treated by embospheres and PVA alone were significant; otherwise there was no significant difference between patients treated with only one embolic material (embospheres and PVA) versus those treated with other additional embolic agents in each group. The patients treated with microspheres had a minor quantification of intraoperative blood loss compared to those who received PVA particles. Furthermore, they had a minor increase of bleeding related to the delay time between embolization and surgery. The use of additional embolic material did not improve the efficacy of the procedure in either group of patients.