For taxation systems, incremental technological developments and adoption have led to a need to re-examine drivers of tax (non) compliance behaviour. Looking at the Australian tax compliance system, technological tools include for example data matching toolkit, pre-filling data and electronic tax lodgement known as “myTax”. In this paper, we develop a new model, labelled Taxpayer Propensity to Offend (TPtO), to explain taxpayers’ intentional tax non-compliance behaviour and to account for possible variation of the drivers of tax compliance between two groups of taxpayers: those lodging using myTax and those lodging via a tax agent. In doing so, a sample of non-compliant behaviours (or offences) were examined, from not lodging a tax return to failing to pay tax debts. Following the development of a novel vignette-based survey instrument, we collected 422 survey responses from a sample of Australian taxpayers. Data was analysed using Structural Equation Modelling (SEM) and other statistical tests to account for the differences between the two groups of taxpayers. This research revealed that while taxpayers’ propensity to offend is influenced by their perceived severity and perceived control to prevent punishment, other drivers of the TPtO differ across the two groups of taxpayers. This study found a divergence between the degree of how each tax offence is perceived based on the scenario of each construct of the TPtO. It also confirmed that technological infrastructure, such as the data matching toolkit, is positively shaping taxpayers’ compliance behaviour. This study contributes to the literature on tax compliance and technology adoption. Such empirical data can be useful to tax authorities and regulators to gain independent evidence on drivers of taxpayers’ non-compliance behaviour.