Purpose: The purpose of this paper is to report innovative research aimed at ascertaining whether standards for school leaders could be applied to the process of selecting senior secondary school principals for appointment. Specifically, psychometrically robust measures of performance are sought that would sufficiently differentiate performance to support judgement about whether or not to appoint candidates in the public educational authority in Western Australia. Design/methodology/approach: Initially, the authors describe previous research in the application of standards to selection in education, followed by a brief review of selection practices undertaken in two international settings and the Western Australian educational setting. The authors then describe the innovative design of performance-based assessment tasks for the selection process including task and rubric development, rater training and data validation. The Rasch measurement model is used to analyse the datasets gathered during three iterations of selection process for the Western Australian education authority. Findings: The Rasch analysis of each dataset provides evidence of construct validity and a robust measure of reliability. The person and item location distributions indicate the tasks were better targeted for the highest performing candidates in the second and third iterations and that fine-grained discrimination was evident across the candidate locations. Originality/value: The paper describes psychometrically robust measures of performance that would sufficiently differentiate performance to support judgement about whether or not to appoint candidates in the public educational authority in Western Australia. Specifically, the paper describes the development of performance-based tasks over three years and demonstrates the application of Rasch analysis to examine the psychometric properties of the assessment tasks. (Contains 3 tables and 4 figures.)