Twenty-five new lead isotope data are provided for 23 sulfide and 2 oxide minerals from 20 ore deposits embracing the wide variety of metallic mineralization present in the Cordillera de Carabaya, southeastern Peru, segment of the central Andean Inner Arc domain. The deposits are subdivided into the following groups: (1) Cu + or - W + or - Mo + or - Sn veins hosted by Permian and Triassic, metaluminous to peraluminous granitoid plutons; (2) Pb-Zn-Ag-Cu veins spatially related to small, Upper Cretaceous stocks of granodioritic composition; (3) major Sn-Cu-Pb-Zn-Ag lodes associated with strongly peraluminous cordierite-biotite monzogranite stocks of mid-Tertiary age; (4) Sn, W, Cu, Pb, Zn, Ag, and Sb veins of undetermined but probably mid-Tertiary age; and (5) Fe + or - Mn + or - Ba manto deposits of poorly constrained age.Collectively, the isotopic data indicate a dominantly crustal lead component in all suites, with analyses plotting well above the 206 Pb/ 204 Pb- 207 Pb/ 204 Pb orogen growth curve of Zartman and Doe (1981); for several mineralized centers (Permo-Triassic and a probable mid-Tertiary Sb deposit), the data overlap the upper crust curve. Although lead isotope data for Cretaceous and mid-Tertiary (group 3) ores exhibit tight clustering in all Pb-Pb plots, the data for Permo-Triassic ores display a large spread and represent the most radiogenic compositions, with 206 Pb/ 204 Pb values ranging from 18.819 to 25.18. Isotopic analyses of acid leaches of sulfide ores from two mineralized centers of Permo-Triassic age define regression lines in 206 Pb/ 204 Pb- 207 Pb/ 204 Pb space and, in at least one case, indicate the presence of a 1,700-Ma radiogenic lead component. Data for group 4 mineralization, of probable mid-Tertiary age, plot near group 3 data in all Pb-Pb diagrams, providing some corrobofating evidence for their inferred mid-Tertiary age; the only exception is the Sb mineralization, which is relatively enriched in both 207 Pb and 208 Pb.The isotopic compositions cannot be accommodated by a single reservoir model because (1) the implied age difference (ca. 100-125 Ma) between groups 2 and 3 in Pb-Pb plots is greater than that for the host granites (ca. 55 Ma); and (2) group 1 mineralization, the oldest analyzed, contains the most radiogenic lead. Multiple source regions or complex mixing models are therefore predicated for the region as a whole.A comparison of the southeastern Peru ore lead data with those available for the metallo-genically similar northwestern Bolivian Sn-W-Ag province reveals broadly similar results for Tertiary Sn polymetallic ores but radically different compositions for (Permo-)Triassic mineralization. Thus, different source regions or processes are indicated for at least the older mineralization. In addition, whereas the Mesozoic and Cenozoic Bolivian ore lead compositions could be accommodated by a common evolving source reservoir(s) (Tilton et al., 1981), the Peruvian data cannot.The lead isotope compositions reported here overlap extensively with much of the Pb isotope data available for central Andean Main Arc mineralization, although our results lie toward the more radiogenic ends of defined fields. We therefore conclude that a model involving mixing of leads from a single source (mantle?) with various crustal reservoirs can account for the observed data spread in Pb-Pb plots. This is consistent with the petrochemistry of the Inner Arc granitoid igneous suites which indicates involvement of mantle-derived components in the anatectic granitoid rocks via magma mixing processes. An important implication of this study is that multiple reservoirs have been involved in the metallogenic evolution of the Central Andean tin belt, a conclusion arrived at earlier from petrologic studies of the associated igneous suites.