Recently, dynamic norms have been introduced as a motivator for pro-environmental behaviour by highlighting that the prevalence or approval of a behaviour is changing. However, operationalisations of dynamic norms vary considerably across studies. Perhaps as a consequence, the impact of dynamic norms on pro-environmental behaviour is inconsistent. This meta-analysis (79 studies from 49 articles) investigates the overall effect of dynamic norms on pro-environmental behaviour, and whether differences in operationalisations affect dynamic norms’ effects. For this, we identify five key elements of dynamic norms: (1) the reference group, (2) the (pro-environmental) behaviour performed or approved by the reference group, (3) the direction of change, (4) the subject of change, and (5) the strength of change (i.e., rate and persistence of change). Results show that while overall effects of dynamic norms on pro-environmental behaviour were very small (r = .03 – .10), effects indeed differed between operationalisations of these elements. Dynamic norms were more effective when they referred to general reference groups rather than to specific ones, used proscriptive rather than prescriptive messages (i.e., the direction of change), emphasised people, or a combination of people and behaviour, rather than behaviour alone as the subject of change, and when framed over shorter rather than longer timeframes (i.e., a proxy for strength of change). Further, dynamic norms combining descriptive and injunctive appeals were more effective than those relying on a descriptive appeal alone. Recognising key elements of dynamic norms advances theoretical understanding of dynamic norms, informs future research, and provides crucial insights into their practical applicability.