We discuss a property of distance-based measures that has not been addressed with regard to evaluating the geographic concentration of economic activities. The article focuses on the choice between a probability density function of point-pair distances or a cumulative function. We begin by introducing a new cumulative function, M, for evaluating the relative geographic concentration and the co-location of industries in a non-homogeneous spatial framework. Secondly, some rigorous comparisons are made with the leading probability density function of Duranton and Overman (2005), Kd. The merits of the simultaneous use of Kd and M is proved, underlining the complementary nature of the results they provide.