We construct a Hertzsprung-Russell diagram for the stellar/substellar boundary based on a sample of 63 objects ranging in spectral type from M6V to L4. We report newly observed VRI photometry for all 63 objects and new trigonometric parallaxes for 37 objects. The remaining 26 objects have trigonometric parallaxes from the literature. We combine our optical photometry and trigonometric parallaxes with 2MASS and WISE photometry and employ a novel spectral energy distribution fitting algorithm to determine effective temperatures, bolometric luminosities, and radii. Our uncertainties range from ~20K to ~150K in temperature, ~0.01 to ~0.06 in log (L/L_{sun}_) and ~3% to ~10% in radius. We check our methodology by comparing our calculated radii to radii directly measured via long baseline optical interferometry. We find evidence for the local minimum in the radius-temperature and radius-luminosity trends that signals the end of the stellar main sequence and the start of the brown dwarf sequence at T_eff_~2075K, log(L/L_{sun}_)~-3.9, and (R/R_{sun}_)~0.086. The existence of this local minimum is predicted by evolutionary models, but at temperatures ~400K cooler. The minimum radius happens near the locus of 2MASS J0523-1403, an L2.5 dwarf with V-K=9.42. We make qualitative arguments as to why the effects of the recent revision in solar abundances accounts for the discrepancy between our findings and the evolutionary models. We also report new color-absolute magnitude relations for optical and infrared colors which are useful for estimating photometric distances. We study the optical variability of all 63 targets and find an overall variability fraction of 36_-7_^+9^% at a threshold of 15mmag in the I band, which is in agreement with previous studies.