Large metaphyseal bone defects created after benign bone tumor curettage are usually treated by solid bone cement (PMMA) implantation. Porous and absorbable cements are expected to improve clinical results by lowering incidence of aseptic loosening, heat promoted bone necrosis, bone/cement mechanical discrepancies and absence of bone integration. Effervescent components like sodium bicarbonate and citric acid can be mixed to PMMA producing porous cement that fits intraoperative requirements. Furthermore, castor oil polyurethane is a well-known absorbable cement that can be mixed to effervescent components to produce porous and absorbable bone cement. In vivo and in vitro toxicity of both products have never been tested to support its clinical use. The spurious production of toxic elements and the effect of citrate, an anticoagulant byproduct of the effervescent components reaction, were analyzed. In vitro experiments: The three experimental groups consisted of PMMA or one of the commercially available castor oil polyurethane cements (Poliquil ® or Bioosteo ®) mixed to the effervescent components specimens. These groups were compared to classic PMMA solid cement specimens. All the elements produced during preparation and incubation of the cements were analyzed by High Efficiency Liquid Chromatography coupled to Mass Spectrometry and by Resazurine microplate assay of NIH/3T3 and MRC-5 fibroblasts strains culture. In vivo experiments: Femoral defects were created in six rabbits per group and filled with PMMA or Poliquil ® porous cements or left empty. Magnetic resonance of the limb, histology of the limb, kidneys and liver, and coagulation parameters of blood samples collected immediately after and 1, 4 and 7 days after surgery were analyzed. Castor oil polyurethane of both brands (Bioósteo ® and Poliquil ®) released 4,4'-diaminodiphenylmethane. The Poliquil ® brand released 4,4'-diphenylmethane diisocyanate too. Both compounds are considered toxic. The MRC-5 and NIH3T3 cell strains proliferation was decreased in less than 80% in contact with polyurethane cements and less than 65% with PMMA, compared to the control group. Dilution of the medium diminished this effect. We believe that the acidity of the medium due to incomplete reaction of the effervescent components may be the cause of this finding. Liver and kidneys histology showed some slight changes in the same proportion as the control group. We infer that the cements do not cause acute systemic toxicity. In all groups, histologic analysis of the operated area showed small hematomas and a slight foreign body reaction. Significant 12 inflammatory reaction could not be found in any of the study groups. Citrate formation from the effervescent components reaction seems to not interfere with coagulation. Inflammatory reaction around porous cements is similar to that of classic solid PMMA cement. The cytotoxic effect observed in vitro, could not be detected in vivo. Coagulation parameters did not changed at any time after surgery. Magnetic resonance imaging evaluatio...