Global food system governance increasingly relies on multistakeholder initiatives (MSIs) that aim to include those who are affected by and/or affected by an issue. Multistakeholderism’s perceived legitimacy is premised on both its outcomes (output legitimacy) and processes (input legitimacy), the latter in turn based on four key rationales: inclusiveness, procedural fairness, consensual orientation and transparency. To date, evidence on the ineffectiveness of MSI’s outcomes undermines its claims to output legitimacy. While individual case study assessments have also raised concerns over their processes, documenting instances of power asymmetries and corporate capture, there has hitherto been no comprehensive assessment of the input legitimacy of multistakeholderism. This work addresses that gap through interviews with 31 participants working either in or on MSIs. Participants noted significant challenges related to input legitimacy, including that (i) inclusion was often based on pre-existing networks of an MSI’s founders—most of whom were based in the global North—and risked excluding less well-resourced or marginalized actors; (ii) pre-existing power imbalances, both internal and external to the MSI, considerably influenced its processes and structures; (iii) goal-setting was complicated by conflicts of interest and (iv) reliance on informal processes limited transparency. The similarities in challenges across MSIs indicate that these are not attributable to shortcomings of individual MSIs but are instead indicative of wider system constraints. Rather than rely on multistakeholderism as a ‘good’ governance norm, our findings add to evidence that MSIs do not meet output legitimacy and signal that the legitimacy of MSIs in their current form should be questioned.