The formation of planets in our solar system encompassed various stages of accretion of planetesimals that formed in the protoplanetary disk within the first few million years at different distances to the sun. Their chemical diversity is reflected by compositionally variable meteorite groups from different parent bodies. There is general consensus that their formation location is roughly constrained by a dichotomy of nucleosynthetic isotope anomalies, relating carbonaceous (C) meteorite parent bodies to the outer protoplanetary disk and the non-carbonaceous (NC) parent bodies to an origin closer to the sun. It is a common idea, that in these inner parts of the protoplanetary disks, planetesimal accretion processes were faster. Testing such scenarios requires constraining formation ages of meteorite parent bodies. Although isotopic age dating can yield precise formation ages of individual mineral constituents of meteorites, such ages frequently represent mineral cooling ages that can postdate planetesimal formation by millions or tens of millions of years, depending on the cooling history of individual planetesimals at different depths. Nevertheless, such cooling ages provide a detailed thermal history which can be fitted by thermal evolution models that constrain the formation age of individual parent bodies. Here we apply state-of-the-art thermal evolution models to constrain planetesimal formation times particular in the outer solar system formation region of C meteorites. We infer a temporally distributed accretion of various parent bodies from $$<0.6$$
<
0.6
Ma to $$\approx 4$$
≈
4
Ma after solar system formation, with 3.7 Ma and $$2.5-2.75$$
2.5
-
2.75
Ma for the parent bodies of CR1-3 chondrites and the Flensburg carbonaceous chondrite, and $$<0.6$$
<
0.6
and $$<0.7$$
<
0.7
Ma for the parent bodies of differentiated achondrites NWA 6704 and NWA 011, respectively. This implies that accretion processes in the C reservoir started as early as in the NC reservoir and were operating throughout typical protoplanetary disk lifetimes, thereby producing differentiated parent bodies with carbonaceous compositions in addition to undifferentiated C chondrite parent bodies. The accretion times correlate inversely with the degree of the meteorites’ alteration, metamorphism, or differentiation. The accretion times for the CM, CI, Ryugu, and Tafassite parent bodies of 3.8 Ma, 3.8 Ma, $$1-3$$
1
-
3
Ma, and 1.1 Ma, respectively, fit well into this correlation in agreement with the thermal and alteration conditions suggested by these meteorites. Our results indicate that individual planetesimals formed rapidly (i.e., within $$<1$$
<
1
Ma), however, distinct planetesimals formed recurrently throughout the total lifetime of the protoplanetary disk. Rapid individual formation is consistent with streaming instabilities assisted by gravitational collapse. However, obviously not the total dust inventory was consumed at early disk evolution stages, so there must have been some delay mechanisms, e.g. collisional destruction of precursor aggregates due to high impact velocities induced by radial drift phenomena. This counterbalance enabled late ($$>2-3$$
>
2
-
3
Ma) accretion of C planetesimals beyond the snow line which escaped severe planetesimal heating and volatile loss, hence, preserving their volatiles, especially water. Only this delayed formation of water-rich planetesimals allowed Earth to accrete sufficient water to become a habitable planet, preventing it from being a bone dry planet.