From July 2016 to June 2017, the market research company GfK SE introduced the first representative diary study on food waste in German private households. Almost 7,000 households recorded and reported about their handling of food waste (FW) over a period of 14 days each. The study distinguished between avoidable and unavoidable FW and leftovers. The present report introduces the methodology and the findings of that study and also provides comments and recommendations by Thünen Institute (German Federal Research Institute for Rural Areas, Forestry and Fisheries) for future replications from a scientific point of view. The GfK study provides many new insights into how FW is handled in private households in Germany and offers basic knowledge for developing new policy options for reducing food waste. In particular, the type and composition of the discarded foods provides detailed indications of key issues, e.g. 16 % of avoidable FW is cooked / prepared (with a very high energy input). In some cases well-known reasons for disposal are confirmed, and classified in their importance for the topic, such as the result related to the best-before date (BBD): Only 5.8 % of respondents indicate this as a disposal reason. Overall, 57.6 % of food is thrown away due to ´problems related to durability´, 36.6 % were spoiled. When interpreting the results one should bear in mind that carryover effects can obscure the root causes of food disposal, e.g. suboptimal shopping behaviour causes problems regarding durability. Thus, the most important disposal reasons could be reduced by more planned and needs-based shopping and proper storage and processing at home. But this would take much more time, which may seem disproportionate to many consumers. It should be underlined that the extrapolated values for the total food waste of just under 4.4 million tons are to be considered as the lower threshold. About 44 % of this food waste was considered avoidable by the surveyed households. Thus, there is a theoretical potential for improvement, which could be specifically addressed by using additional information from more detailed surveys.