Studies on ultra-endurance suggest that during the races, athletes typically experience three vitality states (i.e., preservation, loss, and revival) at the phenomenological level. Nevertheless, how these states contribute to the management and outcome of performance remains unclear. The aim of this study was to determine whether and how the vitality states experienced by runners and their evolution during a trail race can be used to distinguish finishers from withdrawers. From an enactive and phenomenological framework, we processed enactive interviews and blog posts of race narratives. We distinguished units of meaning, which were grouped into sequences of experience; each sequence was then categorized as one of the three vitality states: state of vitality preservation (SVP), state of vitality loss (SVL) or state of vitality revival (SVR). We analyzed the distribution of these vitality states and their temporal organization at the beginning, in the second and third quarters, and at the end of the races, and we qualitatively characterized runners' adaptations to SVL. Results showed that finishers completed the race in SVP, with overall significantly more sequences in SVP and significantly fewer sequences in SVL than withdrawers. SVR did not discriminate finishers from withdrawers. The temporal organization of the vitality states showed a significant difference in the emergence of SVP from the second quarter of the race, as well as a significant difference in the emergence of SVL from the third quarter of the race. The analysis of adaptations to SVL confirmed that finishers were more capable of exiting SVL by enacting a preservation world when they felt physical or psychological alerts, whereas withdrawers remained in SVL. Our results showed that finishers and withdrawers did not enact the same phenomenological worlds in the race situation, especially in the organization of vitality adaptations and their relationships to difficulties; the cumulative effect of the succession of experienced vitality states differed, as well.