2-Deoxy-D-ribose-5-phosphate aldolase (DERA) belongs to the family of lyases and can form C-C bonds to generate multiple chiral centers, thus providing an interesting route to produce key chiral compounds. However, several problems, such as low activity and poor stability (poor tolerance to high aldehyde concentrations), limit the practical application of DERA in large-scale production. Many approaches have been introduced to address these issues. Specifically, in the last decade, many new DERAs, with high catalytic activity or excellent aldehyde tolerance, have been cloned from various extremophilic microorganisms. In addition, on the basis of the analysis of the catalytic mechanism of DERA, rational design engineering and computational design have been used to reconstruct DERA enzymes to alter their stability and catalytic activity.