Optimization of the Mo-V-Sb mixed-oxide system for the selective oxidation of isobutane to methacrolein by true combinatorial methods primarily is intended to reduce the number of experiments in a broad parameter space. Therefore, an evolutionary approach based on a genetic algorithm has been chosen to screen three generations of 30 catalysts. With the help of automated sol-gel synthesis techniques, a high-throughput continuous flow reactor (16UPCFR), and appropriate software for experimental design, a new catalyst composition with improved performance has been obtained. Finally, the best catalysts were scaled-up to,grain quantities and tested in a continuous-flow reactor unit that was equipped with four parallel reactors (4UPCFR). The final catalyst showed a significantly higher selectivity toward methacrolein at the same isobutane conversion, compared to the initial Mo8V2Sb90Ox catalyst.