Food that is either wasted or lost, rather than being eaten, accounts for around a third of global food production and is linked to several environmental, economic and social issues. The reliable quantification of this wasted food is essential to monitor progress towards the United Nations’ Sustainable Development Goal 12.3, which covers food loss and waste. Currently quantification of food waste is made difficult by many differing definitions, some of which require categorisation of food items into those parts considered edible and those considered inedible. Edibility is difficult to define as it is affected by cultural and social influences. This study presents a novel, easily-replicable, questionnaire-based methodology to categorise ‘borderline’ food items thrown away from households, e.g. parsnip skin, apple cores. The methodology captures self-reported information on what people eat (self-reported consumption) and their perceptions of edibility. Our results for the United Kingdom indicate that, for a given food ‘part’, there is divergence between individuals’ responses to the survey questions: e.g. many people would ‘never’ eat carrot skins, whilst many others would ‘always’ eat them. Furthermore, there is a systematic difference between people’s self-reported consumption and their perceptions of edibility. We suggest that both need to be considered to create a balanced categorisation of edible and inedible parts; we propose a method for incorporating both elements. Within this method, a threshold needs to be applied and the resultant classification, especially of those items close to this threshold, will inevitably be contentious. Despite this, the categorisation of what is considered edible using this methodology reflects the views of the majority of the population, facilitating the quantification of food waste. In addition, we envisage this methodology can be used to compare geographical differences and track changes over time with regard to edibility.