Five endpoints have been chosen to develop new In Silico (QSAR) models. A list of endpoints requiring In Vivo experiments within REACH has been compiled (Van der Jagt et al., 2004) ordered by the number of animals used in the toxicity experiments. Within CAESAR, five endpoints have been chosen among the most "animal-consuming", considering the possibility to develop suitable model for it. If for a given endpoint no suitable model can be produced, the following one in the list has been considered. In this way the QSAR model can be more effective in terms of animal and cost reduction (the two aspects are related).

Preliminary results for each endpoint are accessible below.