We devulcanized ground tire rubber (GTR) with microwaves at different temperatures (140−200 °C) and heating rates (2−18 °C/min). We measured the soluble content and the cross-link density of the samples, and then evaluated them by Horikx's analysis and determined the specific microwave energy during devulcanization and the selectivity parameter of the treatments. We also introduced the K•D number that can be calculated as the product of the selectivity parameter (K) and the relative decrease in the cross-link density (D). The results showed that GTR devulcanized at lower temperatures, selectivity was very high, and the degree of devulcanization was adequate. At higher temperatures (180−200 °C), the degree of devulcanization reached 85%, but the selectivity parameter was low, which indicates severe degradation at this temperature range. At lower temperatures (140−160 °C) and decreasing heating rate, i.e., increasing residence time, the degree of devulcanization showed an increase, but at higher temperatures, the heating rate had no significant effect on the degree of devulcanization. We introduced different devulcanized GTRs (dGTRs) to natural rubber-based mixtures and produced samples with two different mixing methods: conventional mixing and so-called two-step mixing. In the latter one, in the first mixing step, we added vulcanizing agents to the dGTR, and in the second mixing step, we added this mixture to the reference mixture. Two-step mixing helped recover mechanical properties, especially tear strength. Scanning electron microscopy− energy-dispersive spectroscopy and atomic force microscopy showed that microwave devulcanization and the vulcanizing agents mixed with dGTR acted in the cross-section of dGTR.