We have recently compiled a database of the properties of 192 impact craters, which supercedes previous compilations. Using our database, the impact structures found in North America, Europe and Australia have been examined; these cratonic areas have been relatively stable for considerably long geological periods, and thus have been best preserved. It is confirmed that there is a close correlation between the geological epoch boundaries, the epochs of mass extinctions, antral the "timing" of impacts. In addition, the terrestrial cumulative flux of objects >20km is found to be $1.77{ imes}10^{-15}km^{-2}yr^{-1}$, over the last 120 Myr, which is much smaller than the published values in McEwen et al. (1997) and Shoemaker (1998) ($5.6{pm}2.8{ imes}10^{-15}km^{-2}yr^{-1}$. For terrestrial impact structures with D> 50 km, the apparent cumulative flux over the last 2450 Myr is ~50 times smaller than the corresponding value for the Moon. If we assume that the Earth and the Moon suffered the same level of bombardment over this time, this would mean that the actual flux of impacting bodies, capable of making craters with D)50 km, was ~ 50 times larger than the apparent flux estimated from the currently known terrestrial records.