Even though supporting affected populations is the raison d’etre for humanitarian action, organisations are frequently not accountable to populations they aim to serve. Despite several reform movements, the consultation and participation elements of accountability to affected populations (AAP), in particular, are still lacking. Specifically, during the design phases of projects, such involvement is critical since major decisions around implementation are taken. Given that donors usually use their power over NGOs to set requirements around financial and results-based management, they could do the same for AAP and make funding conditional on meeting requirements. The puzzle of this research is (Q1) to what extent do donors try to enhance AAP at the project design stage by setting AAP requirements and (Q2) what motivates donors to set these requirements and given many do not set them, what holds them back. In total, nine (n=9) problem-centred expert interviews were led with donor and NGO representatives, which were complemented by an analysis of n=14 donor documents related to the project design phase. During the data collection and analysis process, it became apparent that a sole focus on top-down approaches through requirements would leave out relevant constraints around AAP requirements in project design phases. Thus, driven by the interview data, two additional questions were included: (Q3) What are possible reasons why the requirements set by donors are not successful in improving AAP practices? (Q4) How could these constraints be overcome?  For Q1, it was found that there is no systemic inclusion of AAP requirements in project design phases by donors. Though there are positive examples and donors emphasize their engagement, NGO representatives shared the view that there is no real push through requirements for the two more complex elements of AAP, namely consultation and participation. The main bottleneck for donors to set stricter requirements were competing priorities, while their main motivation to do so apart from intrinsic motivations was found to be past failure in combination with hope for effectiveness and efficiency gains (Q2). The additional constraints identified as hampering successful implementation of such requirements were NGO, discursive and systemic constraints, which dealt with resource scarcity, process constraints, conceptual unclarity and misunderstandings as well as constraints related to the humanitarian context and the delivery of aid through projects (Q3). Solutions presented by interviewees underline the importance of updated funding procedures, establishing clarity around the concept, and advancing cash-based programming and the localization agenda (Q4). Finally, the grounded theory developed from the interview and analysis process explains the limited success of donor AAP requirements in the following way: First, such requirements are not established in the first place if hindered by donor bottlenecks. If motivations are stronger than bottlenecks, such requirements do not automatically lead to the implementation of meaningful AAP practices, since the requirements are not adapted to the wider context. Donors (1) do not take NGO constraints into account, (2) do not fully realize discursive constraints and (3) only take limited action against systemic constraints. When setting requirements, donors need to inform their strategies by considering these constraints in setting meaningful requirements.