Previous example-based point generation techniques assumed isotropic point distributions and only measured the spatial-distances between points, ignoring a distribution’s directional properties. We will consider anisotropy by accounting for pairwise correlations across multiple directions using the Pair Correlation Function (PCF). Our example-based method will analyze an example distribution and extract pair correlations along each canonical direction to reproduce visually-comparable distributions with matching spatial-statistics. Our anisotropic model is able to produce much more perceptually-convincing results for anisotropic distributions compared to previous methods without losing the ability to handle isotropic point patterns. Several applications include but are not limited to example-based texture synthesis and object placement.