Data ambiguity and invalidity can cause significant expensive issues in the cadastral domain (e.g. legal disputes). An automated data validation can significantly help to reduce the potential issues. Quality assurance has been comprehensively investigated in various domains, however, the validation of 3D cadastral data is still in its early development. The availability of various regular and irregular shapes for 3D cadastral objects and modern building designs has resulted in a critical need for developing validation rules to ensure data validity and quality.The land registry in Victoria, Australia, is investigating the technical requirements for implementing a 3D digital cadastre. The study of 3D cadastral data validation requirements has been part of this ongoing investigation. This study is being undertaken in three main phases including 1) developing geometrical validation rules, 2) developing non-geometrical validation rules, 3) implementing an online service to validate 3D cadastral data. This paper aims to discuss the initial outcomes of the first phase of the aforementioned study which has focused on developing geometrical validation rules for 3D cadastral objects. The paper reviews the development of four geometrical validation rules which have been formalised using mathematical expressions to check the individual 3D parcels and their relationships with adjoining or neighbouring parcels. The first validation rule checks the compatibility of the cancelled parcel against the created parcels. The second rule deals with parcel collision detection which is required for flagging unacceptable intersection of 3D objects. The third rule ensures the faces forming a 3D parcel are flat. The fourth validation rule assures 3D objects are watertight. The paper concludes with a discussion around the impacts of the proposed validation checks on the subdivision process and future research for the Victorian 3D digital cadastre.