Here we present a methodology that quantifies fracture arrangement in space as a function of position (barycenters), size (trace length in 2D), and orientation (variation in strike). We use point process statistics to analyze distances between barycenters and estimate a confidence interval for randomness that is compared with input data. The methodology can discriminate at least four types of spatial arrangements and identifies preferential associations of distances with fracture length and orientation. We apply our method to two naturally fractured formations and show that it can determine fracture arrangements under a significant statistical confidence.