The metaverse is an emerging convergence of technologies (e.g. virtual reality and blockchains) that enables users to experience mixed/extended realities for a range of legitimate purposes (e.g. gaming, entertainment and education). Unfortunately, the crime and security implications of emerging technologies are often overlooked. To anticipate crimes that the metaverse might facilitate, in this paper we report the findings of a nominal group technique (NGT) study which was informed by a state-of-the-art scoping review of the existing literature. We elicited views from two expert groups: 1) a mixed European sample (with participants from law enforcement, industry, academia and the voluntary sector), and; 2) an international sample of law enforcement stakeholders. A total of 22 crime threats were identified in the existing literature and an additional eight were identified by experts. These included sexual offenses, crimes against the person, crime against property, and financial crimes. Participants were asked to rate these according to their likely harm, frequency, achievability and defeat-ability. Ratings were largely consistent across the two samples, with crimes of a sexual nature (e.g. child sexual abuse material and sexual assault), and against the person (e.g. harassment and hate crime) being rated as presenting the highest future risks (ie being high harm and high frequency). The findings are discussed with the aim of informing approaches to preventing crime in the metaverse.