The Jiaodong Peninsula, located in the eastern margin of the North China Craton, contains numerous world-class lode gold deposits that combined have a proven gold reserve of >5000 t. Despite a large amount of research, the source and evolution of hydrothermal fluids responsible for the gold mineralization in this region remains highly debated. The Linglong gold deposit is one of the largest deposits in the Jiaodong Peninsula. Gold lodes at Linglong are predominantly hosted within the Late Jurassic Linglong granite and structurally controlled by NE-trending faults. They are closely associated with large-scale hydrothermal alteration, which is paragenetically characterized by hydrothermal K-feldspar in pre-ore stage I, and sericite-quartz-pyrite in syn-ore stages II and III. The syn-ore alteration mineral assemblage is distributed within and around the gold lodes, whereas the K-feldspar alteration remains mainly distal to the gold lodes. From the unaltered to progressively intensifying K-feldspar-altered Linglong granites, there are decreasing trends for both whole-rock Li concentrations and δ7Li values, which are attributed to Rayleigh fractionation. The two values increase covariantly within the sericite-quartz-pyrite alteration zone towards the mineralization center, suggesting that further fluid-rock interaction and buffering of the ore fluid accounts for the observed Li elemental and isotopic variations. Fluid inclusion assemblages in quartz of stages I-III have homogenization temperatures of 349°-312 °C, 318°-272 °C, and 307°-246 °C, respectively. In-situ oxygen isotopic analysis on stage I quartz reveals δ18OQ values of 10.9–14.7 ‰, corresponding to δ18OW values of 4.3–9.2 ‰ for the equilibrated fluid. The δ18OW values of the equilibrated fluids of stages II and III are 5.6–8.9 ‰ and 5.2–8.8 ‰, respectively. The oxygen isotopic compositions indicate that magmatic fluid is prevailing for the K-feldspar and sericite-quartz-pyrite alterations. A binary mixing model using Li concentrations and δ7Li values further implies that the hydrothermal fluid was mainly sourced from coeval mafic magmas derived from the metasomatized lithospheric mantle. Our study highlights that lithium and in-situ oxygen isotopes can be used effectively to fingerprint the source of hydrothermal fluids and fluid-rock interactions.