Empirical evidence does not fully support the universal nature of the positive interspecific abundance—distribution relationship. We have earlier documented a negative relationship for butterfly species in Finland, but recently our view was again challenged using a small subset of Finnish butterflies as apparent evidence. Here we scrutinize the critique and identify some general conceptual challenges in analyses of interspecific abundance—distribution relationships. We identify the common problem that the abundance—distribution studies include only a small subset of species, and thus reveal only sample characteristics, not overall patterns in complete assemblages. Small subsets of species are also unlikely to have sufficient power to reveal nonlinear relationships. Second, varying definitions of abundance, especially the practice of using a single point estimate to describe average density, further spur the empirical evidence for the abundance—distribution relationship. To get theoretically relevant results abundance and distribution must be defined and operationalised consistently; otherwise macroecology will reduce to simple documentation of ambiguous patterns and aid little in understanding the biological world.