Non-technical summary : Agriculture has been dominated by annual plants, such as all cereals and oilseeds, since the very beginning of civilization over 10,000 years ago. Annual plants are planted and uprooted every year which results in severe disturbance of the soil and disrupts ecosystem services. Science has shown that it is possible to domesticate completely new perennial grain crops, i.e. planted once and harvested year after year. Such crops would solve many of the problems of agriculture, but their development and uptake would be at odds with the current agricultural technology industry. Technical summary : Agriculture is arguably the most environmentally destructive innovation in human history. A root cause is the reliance on annual crops requiring uprooting and restarting every season. Most environmental predicaments of agriculture can be attributed to the use of annuals, as well as many social, political, and economic ones. Advances in domestication and breeding of novel perennial grain crops have demonstrated the possibility of a future agricultural shift from annual to perennial crops. Such a change could have many advantages over the current agricultural systems which are to over 80% based on annual crops mainly grown in monocultures. We analyze and review the prospects for such scientific advances to be adopted and scaled to a level where it is pertinent to talk about a perennial revolution. We follow the logic of E.O. Wright's approach of Envisioning Real Utopias by discussing the desirability, viability, and achievability of such a transition. Proceeding from Lakatos' theory of science and Lukes' three dimensions of power, we discuss the obstacles to such a transition. We apply a transition theory lens to formulate four reasons of optimism that a perennial revolution could be imminent within 3–5 decades and conclude with an invitation for research.