One major aspect of innovation by the fragrance industry is the discovery of new materials as fragrance ingredients to impart novel and superior consumer experiences.Fragrance ingredients are widely used in a great variety of both personal care and household products, and in many countries, market introduction of new ingredients requires regulatory approval in advance.Safety assessment is often an integral part of regulatory review, which mandates submission of toxicological and environmental data generated by studies following commonly accepted testing guidelines.Therefore, it is important to identify critical concerns at the early stage of new material discovery to guide further development and gauge certainty of regulatory success.In addition, the development and use of alternative methods also benefit the use of new materials in areas where animal testing is increasingly discouraged.Here we describe a scientific framework established within Givaudan to identify safety "alerts" in our new material discovery program.Based on use and exposure characteristics, it aims at using in silico and in vitro methods to determine whether new material candidates or their metabolites are mutagenic, sensitizing to the skin, acutely toxic to aquatic species, or have the tendency to persist or bioaccumulate in the environment.The initial step is a customized in silico workflow built upon read-across to existing fragrance molecules and several SAR tools based on known structural alerts.This step may identify "red flags", but more importantly, facilitate prioritization of tests that need to be performed.On the human health side, skin sensitization is a key endpoint for topical products including fragrances; and therefore, if deemed a potential risk, will trigger the need to conduct two in vitro methods, namely a cell culture-based method called KeratinoSensTM and a cell-free method called peptide reactivity assay.Both methods examine the potential of a new molecule to react with proteins and consequently elicit immune responses.If systemic toxicities become a concern, an in vitro metabolism study may be conducted with rat primary hepatocytes in suspension to generate metabolic maps.As a follow up, plated hepatocytes can also be used to further characterize physiological effects.In addition, in vitro skin irritation and skin metabolism studies may also be conducted on a need basis.On the environmental side, in order to predict acute toxicity to fish, an in vitro assay measuring toxicity to rainbow trout gill cells is conducted based on the premise that fish toxicity is mainly governed by effects on gill cells.Bioaccumulation is another critical endpoint for which guideline testing requires the use of a large number of animals.We adopt an approach that combines in silico screening and in vitro to in vivo extrapolation (IVIVE) model.The in silico screening uses computer models to estimate BCF based on physical chemical properties such as LogKow.The use of IVIVE is based on the premise that biotransformation reduces bioaccumulation.In vitro metabolism is measured by the rainbow trout liver S9 assay, and the in vitro intrinsic clearance rate is incorporated into the IVIVE model to calculate whole fish metabolism and in sequence predict BCF, refining result from the in silico screening.These methods have been used to assess a number of new fragrance material candidates within Givaudan, and results compared against those of benchmark fragrance materials with data generated from guideline testing in vivo.Information generated by these methods can serve as the basis for preliminary regulatory assessment; and once accumulated for a number of materials, such information can be used to develop structural learnings internally, that can be in turn derived to guide the design for safety by chemists.In the long run, accumulation of data and experience with these methods will increase our scientific confidence, leading to robust safety assessment with less use of animals.