Plants depend on the enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) for CO2 fixation. However, especially in C3 plants, photosynthetic yield is reduced by the formation of 2-phosphoglycolate, a toxic oxygenation product of Rubisco, which needs to be recycled in a high-flux–demanding metabolic process called photorespiration. Canonical photorespiration dissipates energy and causes carbon and nitrogen losses. Reducing photorespiration through carbon-concentrating mechanisms, such as C4 photosynthesis, or bypassing photorespiration through metabolic engineering is expected to improve plant growth and yield. The β-hydroxyaspartate cycle (BHAC) is a recently described microbial pathway that converts glyoxylate, a metabolite of plant photorespiration, into oxaloacetate in a highly efficient carbon-, nitrogen-, and energy-conserving manner. Here, we engineered a functional BHAC in plant peroxisomes to create a photorespiratory bypass that is independent of 3-phosphoglycerate regeneration or decarboxylation of photorespiratory precursors. While efficient oxaloacetate conversion in Arabidopsis thaliana still masks the full potential of the BHAC, nitrogen conservation and accumulation of signature C4 metabolites demonstrate the proof of principle, opening the door to engineering a photorespiration-dependent synthetic carbon–concentrating mechanism in C3 plants. Data analysis was performed in R. For analysis of gas exchange measurements, the “plantecophys” package was used (55). The data are summarized in Datasets S1–S10. All other study data are included in the article and/or supporting information, available at https://doi.org/10.1073/pnas.2022307118 Dataset S1: Enzymatic activity of BHAC enzymes in Arabidopsis rosette leaves. For ISR the rate of percentual 15N label enrichment in aspartate was quantified. Shown mean and standard deviation (SD). Dataset S2: Metabolome of BHAC plants. Shown is mean and standard deviation (SD) of the calculated relative amount per mg fresh weight of four biological replicates per genotype for each condition. Dataset S3: Ammonium quantification in BHAC plants. Shown is mean and standard deviation (SD) for four biological replicates per genotype per condition measured in technical triplicates. Dataset S4: Phenotyping of BHAC plants. Shown is mean and standard deviation (SD) of five biological replicates per genotype per condition. Dataset S5: A/Ci curve measurements of BHAC plants. Shown is mean of four biological replicates per genotype. Dataset S6: Light response measurements of BHAC plants. Shown is mean of four biological replicates per genotype. Dataset S7: Metabolite levels of phosphorylated intermediates and glyoxylate in air-grown plants. Shown is mean and standard deviation of ≥ 3 replicates. Dataset S8: Metabolome of ggt1-1 complementation lines with AGAT. Shown is mean and standard deviation (SD) of four biological replicates. Dataset S9: Enzymatic activity of AGAT and GGT in Arabidopsis rosette leaves of the ggt1-1 complemention lines. Shown mean and standard deviation (SD) of three biological replicates measured in technical triplicates. Dataset S10: O2-Dependency of CCP was measured at 4% O2. Shown is the mean ±SD of n ≥ 3.