Significant advancements in photogrammetric Structure-from-Motion (SfM) software, coupled with improvements in the quality and resolution of smartphone cameras, has made it possible to create ultra-fine resolution three-dimensional models of physical objects using an ordinary smartphone. Here we present an open-source modeling framework for creating three-dimensional models of vegetation structure using smartphone video of a user-defined sampling space (e.g., 1 m2). Our main objective was to evaluate the accuracy of our SfM sampling method in predicting above-ground biomass relative to traditional estimation techniques. A series of permanent 1-m2 quadrat sampling sites at the Jornada Experimental Range in southern New Mexico were used to develop the SfM sampling method, with each quadrat differing in its proportions of shrub vs. herbaceous vegetation composition.  A smartphone mounted on a selfie stick was used to capture video of each 1-m2 quadrat by circumnavigating it while moving the smartphone up and down to capture a range of image angles. Images were extracted from video frames at sampling rate of 4 frames per second and used to create densely reconstructed point clouds using open-source SfM software. Point clouds were georeferenced using a set of 3 control points located on the ends of a 0.25-m3 PVC corner frame placed at each site prior to imaging. Canopy volume was calculated for each point cloud using a foliar canopy approach, and used to model above ground biomass using published linear regression models. These results were compared to traditional nondestructive biomass estimates that were calculated by measuring the dimensions (cover and height) of individual plants or plant parts within each quadrat and converted to aboveground biomass using allometric equations. Preliminary results show a strong correspondence between measured and SfM modeled biomass. Our presentation will describe the details of this method and discuss its potential utility in increasing the accuracy and repeatability of field-based biomass estimation.