The Abitibi Belt is the largest continuous greenstone belt in the Superior Province of the Canadian Shield. It comprises several composite komatiitic–tholeiitic–calc-alkalic and sedimentary sequences that are folded, transected by major faults, and intruded by various generations of plutonic rocks. Precise U–Pb geochronology has been carried out in the belt for the past decade to solve chronostratigraphic and metallogenetic problems. This paper presents new zircon ages and reassesses previously published ones, now refined by the addition of abraded and concordant zircon analyses.Volcanic and subvolcanic units of the Timmins area yield the following ages: 2727 ± 1.5 Ma for a tuff of the upper Deloro Formation; 2703 ± 1.5 Ma for a tuff of the upper Tisdale Formation; and 2698 ± 4 Ma for the Krist fragmental, assigned to the top of the Tisdale Formation. The age of a dunite intrusive into the upper Deloro Formation is revised at 2707 ± 3 Ma, whereas the Aquarius diorite east of Timmins yields a poorly defined age of 2705 ± 10 Ma. In the Lake Abitibi – Kirkland Lake region, the following dates were obtained: 2713 ± 2 Ma for a porphyritic unit of the Hunter Mine Group; 2714 ± 2 Ma for a rhyolite at the base of the mainly tholeiitic to komatiitic Stoughton–Roquemaure Group; 2701 ± 2 Ma for porphyritic rhyolite of the Blake River Group; 2701 ± 2 Ma for a tuff of the Skead Group; and for a pyroclastic unit at the base of the Larder Lake Group. These data are generally consistent with the earlier proposed stratigraphic subdivisions and correlations. However, there are apparent age reversals, for example between the Larder Lake and Skead groups, that could support the concept of thrusting and tectonic thickening to explain particular lithologic relationships and the considerable stratigraphic thickness of the supracrustal sequences in the Abitibi Belt.One part of the study was dedicated to the problem of gold mineralization in the Timmins area. The zircon age of 2690 ± 2 Ma for the Paymaster porphyry, a less well defined but probably identical age for the Preston porphyry, and dates of 2689 ± 1 Ma for the Pearl Lake porphyry, 2691 ± 3 Ma for the Millerton porphyry, and 2688 ± 2 Ma for the Crown porphyry show that these intrusions were formed during a well-defined, short-lived episode unrelated to volcanism; furthermore, a date of for an albitite, which predates Au mineralization, demonstrates that Au is spatially but not genetically related to the porphyries.Finally, two ages are reported for late tectonic potassic intrusions: a refined age of 2678 ± 2 Ma for the Garrison stock east of Matheson and a precise zircon (and titanite) age of 2680 ± 1 Ma for the Otto stock near Kirkland Lake.