Clematis tientaiensis is a wild plant species with an extremely small population endemic to Zhejiang Province, China. Currently, this species faces severe challenges regarding natural population regeneration and is at a high risk of decline and recession. To develop a scientific conservation strategy, this study systematically observed the flowering phenology, pollen viability, stigma receptivity, and flower-visiting insect behavior of C. tientaiensis, and combined these observations with artificial controlled pollination to examine its niche characteristics and breeding system. The results indicated that, in terms of ecological characteristics, the niche breadth of C. tientaiensis was greatest in the shrub and herb layers (8.33). Furthermore, the mean niche overlap among main associated species of C. tientaiensis was consistently lower than 0.20. This low overlap suggests significant independent niche differentiation among species, indicating that C. tientaiensis occupies a distinct ecological space with minimal interspecific competition for resources. The flowering phenology observations revealed that the single-flower flowering period of C. tientaiensis lasted about 9 days, and the group flowering period lasted 28 days. The pollen viability was highest (84.45%±1.79%) at the initial full-flowering stage, while stigma receptivity was strongest during the initial flowering stage. The distinct temporal separation between pollen maturation and maximum stigma receptivity confirms the existence of dichogamy (specifically, temporal isolation of male and female functions) in this species. The investigation into pollinators recorded three species of flower-visiting insects. Among them, Episyrphus balteatus and Halictus tumulorum were identified as effective pollinators, while Nonarthra cyanea acted as a destructive flower visitor, damaging floral organs rather than aiding in reproduction. Artificial pollination experiments demonstrated that there was no significant difference in the seed-setting rate between cross-pollination treatment and the open-pollination control group (46.19%±11.50%). In contrast, seed-setting rates under self-pollination, non-emasculated bagging, and emasculation without bagging were significantly lower (P < 0.05), and no fruit formation occurred in the emasculated bagging treatment. Synthesizing these findings, the breeding system of C. tientaiensis is characterized as predominantly outcrossing (xenogamy) with a strict dependence on insect vectors for pollination. The species exhibits clear evidence of inbreeding depression. The renewal of the population is currently limited by pollination obstacles caused by the sparse spatial distribution of individuals and a scarcity of effective pollinators. Consequently, it is recommended that population recovery be supported through artificial pollination and spatial optimization of population structure. These findings provide a crucial basis for understanding the endangered mechanisms of C. tientaiensis and for formulating targeted scientific conservation measures.