To increase the energy density of lithium-ion batteries (LIBs), high-capacity anodes which alloy with Li ions at a low voltage against Li/Li⁺ have been actively pursued. So far, Si has been studied the most extensively because of its high specific capacity and cost efficiency; however, Ge is an interesting alternative. While the theoretical specific capacity of Ge (1600 mAh g^–1) is only half that of Si, its density is more than twice as high (Ge, 5.3 g cm^–3; Si, 2.33 g cm^–3), and therefore the charge stored per volume is better than that of Si. In addition, Ge has a 400 times higher ionic diffusivity and 4 orders of magnitude higher electronic conductivity compared to Si. However, similarly to Si, Ge needs to be structured in order to manage stresses induced during lithiation and many reports have achieved sufficient areal loadings to be commercially viable. In this work, spinodal decomposition is used to make secondary particles of about 2 μm in diameter that consist of a mixture of ∼30 nm Ge nanoparticles embedded in a carbon matrix. The secondary structure of these germanium–carbon particles allows for specific capacities of over 1100 mAh g⁻¹ and a capacity retention of 91.8% after 100 cycles. Finally, high packing densities of ∼1.67 g cm^–3 are achieved in blended electrodes by creating a bimodal size distribution with natural graphite.