Antibiotic and herbicide resistance genes are the most common marker genes for plant transformation to improve crop yield and food quality. However, there is public concern about the use of resistance marker genes in food crops due to the risk of potential gene flow from transgenic plants to compatible weedy relatives, leading to the possible development of ‘‘superweeds’’ and antibiotic resistance. Several selectable marker genes such asaph, nptII, aaC3, aadA, pat, bar, epsp and gat, which have been synthesized to generate transgenic plants by genetic transformation, have shown some limitations. These marker genes, which confer antibiotic or herbicide resistance and are introduced into crops along with economically valuable genes, have three main problems: selective agents have negative effects on plant cell proliferation and differentiation, uncertainty about the environmental effects of many selectable marker genes, and difficulty in performing recurrent transformations with the same selectable marker to pyramid desired genes. Recently, a simple, novel, and affordable method was presented for plant cells to convert non-metabolizable phosphite (Phi) to an important phosphate (Pi) for developing cells by gene expression encoding a phosphite oxidoreductase (PTXD) enzyme. TheptxDgene, in combination with a selection medium containing Phi as the sole phosphorus (P) source, can serve as an effective and efficient system for selecting transformed cells. The selection system adds nutrients to transgenic plants without potential risks to the environment. TheptxD/Phi system has been shown to be a promising transgenic selection system with several advantages in cost and safety compared to other antibiotic-based selection systems. In this review, we have summarized the development of selection markers for genetic transformation and the potential use of theptxD/Phi scheme as an alternative selection marker system to minimize the future use of antibiotic and herbicide marker genes.