CRISPR/Cas9 has been widely used in engineering Saccharomyces cerevisiae for gene insertion, replacement and deletion due to its simplicity and high efficiency. The selectable markers of CRISPR/Cas9 systems are particularly useful for genome editing and Cas9-plasmids removing in yeast. In our previous research, GAL80 gene has been deleted by the plasmid pML104-mediated CRISPR/Cas9 system in an engineered yeast, in order to eliminate the requirement of galactose supplementation for induction. The maximum artemisinic acid production by engineered S. cerevisiae 1211-2 (740 mg/L) was comparable to that of the parental strain 1211 without galactose induction. Unfortunately, S. cerevisiae 1211-2 was inefficient in the utilization of the carbon source ethanol in the subsequent 50 L pilot fermentation experiment. The artemisinic acid yield in the engineered S. cerevisiae 1211-2 was only 20%-25% compared with that of S. cerevisiae 1211. The mutation of the selection marker URA3 was supposed to affect the growth and artemisinic acid production. A ura3 mutant was successfully restored by a recombinant plasmid pML104-KanMx4-u along with a 90 bp donor DNA, resulting in S. cerevisiae 1211-3. This mutant could grow normally in a fed-batch fermentor with mixed glucose and ethanol feeding, and the final artemisinic acid yield (>20 g/L) was comparable to that of the parental strain S. cerevisiae 1211. In this study, an engineered yeast strain producing artemisinic acid without galactose induction was obtained. More importantly, it was the first report showing that the auxotrophic marker URA3 significantly affected artemisinic acid production in a pilot-scale fermentation with ethanol feeding, which provides a reference for the production of other natural products in yeast chassis.