The major biological crisis that marked the end of the Cretaceous fascinates the geoscientists. To understand it, it becomes more and more essential establishing the accurate chronology of the mechanisms that triggered this major event. The boundary is currently estimated at 65.5 Ma, but the high discrepancy between estimations and low resolution of radiometric dating implies the absolute necessity of another high resolution dating tool. The development of cyclostratigraphy and orbital tuning on Neogene series has resulted in estimations with accuracy never reached before, confirming this discipline as a key approach to refine the geological time scale. Our study realised a high resolution cyclostratigraphic analysis on ODP cores, in order to astronomically calibrate the K-P boundary. Holes 762C, 1001A and 1001B present an excellent recovery, a well constrained magnetostratigraphic frame, and, above all, well defined colour cycles which could be linked to the orbital forcing of the sedimentation. Our analysis was thus applied on the grey-scale variations obtained by scanning photographs of the cores. This proxy required a pre-treatment of cracks, coded in black, and therefore a major source of noise. Variations in the photograph luminance, inducing artificial cycles, have been eliminated too. Spectral analysis and amplitude spectrograms methods allowed to highlight the orbital control of sedimentation of all sites, with a predominance of precession, 400 ky and 100 ky eccentricity. Cycle counting allowed us to estimate durations of magnetochrons C29n to C27n, excepted the C27r. Orbital tuning of grey level variations was realised using the stable 400 ky or 100 ky eccentricity as tie-point. Comparison with the most recent theoretical astronomical solution permitted us to propose an estimation of the K-P boundary age.