It is common practice to manipulate CO2 and O2 mole fraction during gas-exchange experiments to suppress or exacerbate photorespiration, or simply carry out CO2 response curves. In doing so, it is implicitly assumed that metabolic pathways other than carboxylation and oxygenation are altered minimally. In the past few years, targeted metabolic analyses have shown that this assumption is incorrect, with changes in the tricarboxylic acid cycle, anaplerosis (phosphoenolpyruvate carboxylation), and nitrogen or sulphur assimilation. However, this problem has never been tackled systematically using non-targeted analyses to embrace all possible affected metabolic pathways. Here, we exploited combined NMR, GC–MS, and LC–MS data and conducted non-targeted analyses on sunflower leaves sampled at different O2/CO2 ratios in a gas exchange system. The statistical analysis of nearly 4,500 metabolic features not only confirms previous findings on anaplerosis or S assimilation, but also reveals significant changes in branched chain amino acids, phenylpropanoid metabolism, or adenosine turn-over. Noteworthy, all of these pathways involve CO2 assimilation or liberation and thus affect net CO2 exchange. We conclude that manipulating CO2 and O2 mole fraction has a broad effect on metabolism, and this must be taken into account to better understand variations in carboxylation (anaplerotic fixa- tion) or apparent day respiration.