As opposed to a decision tree approach, which provides distinct yes/no decisions leading to a specific next step, the weight-of-evidence approach is more flexible and focuses on the totality of the data in a holistic manner. Ideally, the predictive value of each piece of evidence would be well understood to give certain types of data more weight' than others during the assessment, thus leading to more confi- dence in the overall assessment. Currently, the predictive value of individual data may not be clearly defined, but additional experience with the assays and methods utilized d will continue to improve their predictive value. This weight of evidence approach, in which the individual com ponents of the safety assessment process are prioritized into two tiers of testing, depicted in Figs. 1 and 2, optimizes the efficiency and robustness of the process. Case studies of transgenic protein safety, some derived from peer-reviewed literature and others from regulatory submissions, demon- strate the merits and limitations of the framework. Specif- ically this framework includes: