The label, Participatory Action Research (PAR), seems to be a good one, describing research oriented towards making change in which the interested parties actively participate. In identifying key principles or characteristics, proponents begin to tell us more about what drives it: a collective commitment to participation and democracy at all stages of the research process, from identifying issues to finding useful solutions (McIntyre, 2008; Reason & Bradbury, 2008). This situates PAR as a self-conscious reaction against 'traditional' and particularly positivistic approaches to social scientific researchwhereby 'neutral' researchers are in control of identifying research questions, extracting data from participants and deciding what it meansand points towards its ethical and political coordinates , as both critique and response to inequitable balances of power and resource. As Saija (2014) has made clear in this journal, this is not only political but also ideological, a commitment to engaging in democracy as process and working towards social change. It is hard to find a route to PAR within planning, then, without understanding it to some degree as a turn away from planning's problematic modernist legacy. We are implicated by association with a tradition that planned for people, that gave perceived material improvement with one hand whilst disenfranchising with the other, undoing community cohesion and attachments to place that had developed in some cases over many generations. From this starting point PAR can seem to represent a new moral benchmark, affirming our commitment to social justice and engaging participants from the start of projects in identifying, investigating and finding solutions to problems (McIntyre, 2008). As a way of enacting the laudable values that are still central to the planning project in collaboration with impacted people, we might even characterise this as an attempted return to planning's roots as a social movement. This ethical drive can lead to a temptation, however, to rely heavily on the distinction between good or genuine PAR and PAR that fails to hit the mark. Whilst thinking about what makes good PAR has to be central to reflective practice, it is also the case that any attempt to enact it will end up being, in some sense, an exercise in failure; or perhaps more constructively, in learning from failure. In highlighting this aspect of PAR I hope to contribute constructively to ongoing debates within the discipline and this journal, particularly in bringing together Raynor's (2019) recent contribution and the Interface on learning from mistakes (Campbell, Forester, & Sanyal, 2018). Raynor persuasively argues that Early Career Researchers (ECRs) face particular structural barriers and disproportionate challenges in conducting PAR. My aim is to shed a different light on these issues through offering a complementary perspective, based on my experience as a PhD student and an ECR.