Recent improvements to astronomical modeling of the Solar System have contributed to important refinements of the Cenozoic time scale through astronomical calibration of sedimentary series. We extend this astronomical calibration into the Cretaceous, on the base of the 405 ka orbital eccentricity variation thanks to the recovery of well preserved Maastrichtian sedimentary series from ODP (Ocean Drilling Program) Holes 1258A (Leg 207, Equatorial Atlantic), 1267B (Leg 208, South Atlantic), 762C (Leg 122, Indian Ocean) and DSDP (Deep Sea Drilling Program) Hole 525A (Leg 74, South Atlantic). Cyclostratigraphic analysis is performed on high resolution measurements of magnetic susceptibility on sediments cored during Legs 207 and 208, and gray level variations of sediment color obtained from core photographs from Legs 122 and 74. Astronomical control on sedimentation is evident at every site, with the presence of cycles corresponding to forcing by precession, obliquity and orbital eccentricity variations. Identification of these cycles leads to the definition of a detailed cyclostratigraphic frame covering nearly 8 Ma, from the upper Campanian to the Cretaceous/Paleogene (K/Pg) boundary. Durations of each magnetochron from C32r.2r to C29n are inferred by cycle counting. Astronomical calibrations of Maastrichtian sedimentary series are proposed, based on the 405 ka eccentricity variation according to the most recent astronomical solution La2010a. Two different ages are suggested for the K/Pg boundary, considering the uncertainty of the long-term variation of the 405 ka eccentricity cycle. The first proposal provides a Cretaceous/Paleogene boundary age of 65.59 ± 0.07 Ma and the second an age of 66 ± 0.07 Ma, which is coherent with the most recent radio-isotopic datings. Magnetochron boundaries and the Campanian/Maastrichtian boundary are dated relative to these numerical ages of the K/Pg boundary.