The aim of this research was to assess the combined effects of three heavy metals (copper, lead, cadmium) on the fertilization and offspring quality of the sea urchin Paracentrotus lividus at EC50, NOEL, and EC1 concentrations. The observed data were compared with the predictions derived from approaches of Concentration Addition (CA) and Independent Action (IA) in order to evaluate the proper prediction of the observed mixture toxic effect. The P. lividus embryotoxicity of trace metals decreases as follows: Cu > Pb > Cd at all toxicity concentration tested. EC50 mixture revealed less toxic only than Cu; EC50 was 0.80 (± 0.07) mg/l, the offspring malformations were mainly P1 type (skeletal alterations) up to 20% mixture concentration, and P2 type from 70% concentration. The NOEL and EC1 mixtures evidenced that all compounds contribute to the overall toxicity, even if present at low concentrations: the former EC50 was 0.532 (± 0.058) mg/l and the latter was 1.081 (± 0.240) mg/l. The developmental defects observed were mainly P1 type in both mixtures. Both CA and IA models did not accurately predict mixture toxicity for EC50 and NOEL mixtures. Instead, EC1 mixture effects seemed well represented by the IA model. The protective action of the CA model, although quite accurate when applied to simple biological systems like algae and bacteria, but failed to represent the worst-case in this study with more complex organisms. It would be useful to introduce in the models one or more factors that take into account the complexity of these biological systems.