The ever-expanding repertoire of biocatalysts and advancements in genetic manipulations has led to the customization of natural enzymes and the generation of “new-to-nature” catalytic functions. Computational tools greatly assisted (semi-)rational protein engineering approaches by more reliable predictions of mutations, ultimately, catering for desired traits. Their combination with in vitro compartmentalization and microfluidic techniques elevated the number of protein variants being efficiently screened and selected from a few hundreds to libraries containing hundreds of millions of mutants. This chapter highlights the engineering of proteins from autotrophic organisms including the famous RuBisCO and the adaption of enzymes from heterotrophs to function in autotrophs and vice versa. Together, these examples demonstrate the capability of modern protein engineering techniques to transform autotrophic microorganisms into competitors to wellestablished heterotrophic hosts to tackle today’s challenges like unprecedented enzymatic circuits for the fixation of CO2, the stewardship of resources and the sustainable production of value-added chemicals.