Objective. To develop and characterize novel methods of extreme spatially fractionated kV radiation therapy (including mini-GRID therapy) and to evaluate efficacy in the context of a pre-clinical mouse study.Approach. Spatially fractionated GRIDs were precision-milled from 3 mm thick lead sheets compatible with mounting on a 225 kVp small animal irradiator (X-Rad). Three pencil-beam GRIDs created arrays of 1 mm diameter beams, and three 'bar' GRIDs created 1 × 20 mm rectangular fields. GRIDs projected 20 × 20 mm2fields at isocenter, and beamlets were spaced at 1, 1.25, and 1.5 mm, respectively. Peak-to-valley ratios and dose distributions were evaluated with Gafchromic film. Syngeneic transplant tumors were induced by intramuscular injection of a soft tissue sarcoma cell line into the gastrocnemius muscle of C57BL/6 mice. Tumor-bearing mice were randomized to four groups: unirradiated control, conventional irradiation of entire tumor, GRID therapy, and hemi-irradiation (half-beam block, 50% tumor volume treated). All irradiated mice received a single fraction of 15 Gy.Results. High peak-to-valley ratios were achieved (bar GRIDs: 11.9 ± 0.9, 13.6 ± 0.4, 13.8 ± 0.5; pencil-beam GRIDs: 18.7 ± 0.6, 26.3 ± 1.5, 31.0 ± 3.3). Pencil-beam GRIDs could theoretically spare more intra-tumor immune cells than bar GRIDs, but they treat less tumor tissue (3%-4% versus 19%-23% area receiving 90% prescription, respectively). Bar GRID and hemi-irradiation treatments significantly delayed tumor growth (P < 0.05), but not as much as a conventional treatment (P < 0.001). No significant difference was found in tumor growth delay between GRID and hemi-irradiation.Significance. High peak-to-valley ratios were achieved with kV grids: two-to-five times higher than values reported in literature for MV grids. GRID irradiation and hemi-irradiation delayed tumor growth, but neither was as effective as conventional whole tumor uniform dose treatment. Single fraction GRID therapy could not initiate an anti-cancer immune response strong enough to match conventional RT outcomes, but follow-up studies will evaluate the combination of mini-GRID with immune checkpoint blockade.