This article documents the methodology and results of a risk analysis for seven plays in the frontier More and Voring basins. The play types represent different deep-marine systems and span a complete transect from upper slope to basin floor. Because of the sparse well data coverage typical of frontier basins, the play models are seismically driven, resulting in high-level qualitative risk assessments, dividing the plays into low-, medium-, and high-risk categories. The objectives of the risk assessment were both to develop an understanding of the distribution of promising reservoir thickness, architecture, and net to gross in each play type and, further, to evaluate the range of uncertainties of the depositional model and the quality of the available database. We demonstrate here how the effect of perceptions of risk in constituting a valid play can be evaluated during the ranking process and how the process can be used to systematically screen for future exploration opportunities. The workflow is rooted in a risk analysis of the geological models because this represents an integrated interpretation of the geology with respect to reservoir distribution, architecture, and presence of seal. In conclusion, the prerequisites for proper qualitative risk assessment of geological models are an awareness of reservoir variability and the range of depositional attributes. Play risk consists of regional risk elements, determined by regional depositional models. Equally important is the uncertainty introduced by poor or lacking data. Differentiating between play risk and lack of data improved decision making, i.e., whether to acquire acreage, data, or both.