Determining the number of evolutions of an adaptive novelty is primordial to understand its evolutionary significance. C-4 photosynthesis, an adaptation to low CO2 atmospheric concentration and high temperature, evolved multiple times, but the number of convergent evolutions is still debated. In Poaceae phylogeny, numerous C-4 groups are separated by C-3 taxa, but whether these correspond to independent C-4 origins or a few C-4 evolutions followed by reversals is controversial. The Aristidoideae subfamily is formed by two C-4 genera, Aristida and Stipagrostis, separated by the C-3 genus Sartidia. In the current study, we investigated the evolutionary history of genes encoding phosphoenolpyruvate carboxylases (PEPC) to shed light on the photosynthetic transitions that occurred in Aristidoideae. We identified six distinct PEPC gene lineages that appeared through several rounds of gene duplications before or early during grass diversification. The gene lineage encoding the C-4 PEPC of Stipagrostis differs from those of the other C-4 grasses, including Aristida. These distinct origins of C-4 PEPC genes from these two Aristidoideae genera unequivocally indicate that they integrated the C-4 pathway independently. This highlights the importance of candidate-gene studies when inferring the evolutionary history of a character such as C-4 photosynthesis, one of the greatest evolutionary successes in plant history.