The Tati Greenstone Belt (TGB) in northeastern Botswana hosts numerous Au deposits (Shashe, Mupane and Signal Hill) associated with sulfides, garnet and white mica. In this study, integrative white mica Ar/Ar, garnet U-Pb and sulfides Pb/Pb dating techniques were combined with whole rock and sulfide Pb isotope characteristics to track the source(s) of gold and constrain the timeframes of gold mineralization events. All sulfides and arsenopyrite samples from the TGB yielded overlapping Pb/Pb errorchron ages of 2227 ± 66 Ma and 2220 ± 73 Ma, respectively, which coincide with the Shashe sulfides Pb/Pb errorchron age of 2250 ± 110 Ma. These relatively imprecise Pb/Pb dates suggest heterogeneity in the initial Pb isotope ratios of sulfides. At Mupane, whereas Au mineralization-associated hydrothermal almandine garnet yielded overlapping Tera-Wasserburg lower intercept 206Pb/238U age and a concordia age of 2119 ± 18 Ma (MSWD = 1.03) and of 2105 ± 24 Ma (MSWD of concordance = 1.10), respectively, sulfides produced an errorchron Pb/Pb age of 2873 ± 140 Ma, which despite the large error remains geologically meaningful as it coincides within error with the first Neoarchean Limpopo-Liberian Orogeny (2.70–2.65 Ga), granitoids intrusion emplacement (2.65–2.73 Ga) and deposition of banded iron formation (2.73 ± 0.15 Ga). Ore-related white mica from Signal Hill yielded an overlapping Ar/Ar plateau age of 1987 ± 24 Ma and a weighted mean Ar/Ar age of 1987 ± 13 Ma (n = 15, MSWD = 4.3), which coincide within error with the 2.05–1.95 Ga second Limpopo-Liberian tectonic cycle, herein considered to have possibly triggered the Au mineralization in this area. The obtained radiometric dates point to at least three well constrained gold mineralization events, with two of them possibly triggered by two different regional tectonic events. Lead isotope compositions of most of the sulfides overlap with those of spatially associated schists and granitoids, thus suggesting these units possibly represent Pb and by inference Au rock sources. The genetic model of the TGB Au deposits is consistent with many greenstone-hosted gold deposits worldwide, suggesting our results are not solely of local/regional interest, but can be used to characterize greenstone-hosted gold deposits worldwide.