The idea of ‘subsistence affluence’ was developed in Papua New Guinea (PNG) by an ANU academic in the 1970s to elucidate a situation in which people ate what they grew, and grew enough not to be hungry. That idea has persisted. There is a widespread view that, while there are problems with diet quality, people do not go hungry. In this paper, we challenge that thinking, drawing on several sources, including recent household surveys, and both quantitative and qualitative data. We show that several surveys indicate that significant numbers report themselves to be hungry on a regular basis, and consume calories below the minimum required level. Urban households are more insecure than rural households, but the difference is not large. Reported hunger is widespread. Regression analysis suggests that reported hunger is not randomly distributed, but concentrated where, if it existed, we would expect to find it - that is, among households that are relatively poorer and less educated, not engaged in cash cropping, with fewer household assets, and in disadvantaged districts. We conclude that, whatever the truth in the past, today PNG is, indeed, a hungry country.