Zircon U-Pb geochronology is vital for providing age constraints for the calibration of the geologic timescale. The presence of zircon grains that yield concordant ages younger than the host sedimentary rock is often ignored, attributed to young outliers and possible contaminants. Using in situ U-Pb dating techniques, which analyze zircon crystals drilled directly from polished thin sections, two tuff beds from the Brockman Iron Formation, Pilbara Craton, Australia, yielded weighted mean 207Pb/206Pb ages of 2466 ± 6 Ma and 2469 ± 5 Ma (95% confidence level), consistent with independent age constraints. The tuffs also contain zircons that gave younger, yet concordant, U-Pb dates (2450–2270 Ma) as well as some with highly discordant dates. The “young” zircons have low or negligible common Pb, and display magmatic-type euhedral outlines, oscillatory zonation and Th/U ratios, indistinguishable from syn-depositional zircons, and cannot be readily dismissed on the grounds of high discordance or common-Pb contents. The “young” zircon dates are statistical outliers that overlap with the age of hydrothermal xenotime in deformed, bedding-parallel veins in shale that underlies the iron formation, which yields an age of 2383 ± 4 Ma (95% confidence level). The xenotime in overpressure veins provides independent evidence for hydrothermal fluid flow broadly coeval with isotopic resetting of “young” zircon crystals. Although SEM-CL imaging found no evidence for alteration, element mapping shows that most zircon grains have been altered by fluid infiltration, which proceeded along fractures or radiation-damaged growth zones leading to elevated Fe, Ca and Y contents in the altered domains. Total and partial Pb loss caused by hydrothermal alteration can explain the young concordant and discordant ages, respectively. These results have implications for the use of U-Pb zircon geochronology, particularly in detrital zircon studies, which routinely use the age of the youngest zircon grains to constrain the age of deposition. The erosion and re-deposition or in situ growth of young zircons in low-grade metasedimentary rocks may lead to erroneously young depositional ages.