This is the first time optimism bias, recorded for completed projects, has been used to help provide greater accuracy in the appraisal process. Statistically, the sample of projects in the Mott MacDonald study is necessarily small because, in the time period studied, large public sector procurement was restricted to a relatively limited number of projects. The limited size of the sample is apparent when divided into project types, which do not contain the same number of projects in each category. These limitations have been considered when developing guidance for future appraisals.