The distance of recycling bins from households is often considered important by practitioners, but published evidence for this uses only indirect and self-reported data. This study aims to provide such evidence by obtaining a clean test using measured distances in a walled community with 1200 households with the same building types, local governance, recycling and waste arrangements. The number of deposits each month of food waste for recycling at a designated site are logged via smart-cards allocated per household. The number of days per month that each household deposits showed a highly significant - but small - negative correlation with distance of the bin: fewer householders participate if further away, accounting for 3% of the variation. Surprisingly, there is no variation with distance among those who do participate: their recycling frequency does not vary. This second result is not consistent with the first in terms of cost/benefit concepts assumed by government planners, nor with the static theories of behaviour currently used in waste management research. We recommend that recycling practitioners note the smallness of the contribution of distance to recycling performance, and not overrate it. And we recommend that researchers make better use of non-static models (which model different stages towards behaviour change), which our second result appears to call for.