The Montagnais impact crater is presently the only site in the ocean where the effect of a meteorite fall on marine organisms has been studied. The impact crater is 45 km in diameter and was formed at 50.8 Ma by a fall of probably an old cometary. nucleus 3.4 km in diameter, into shallow (<600 m) ocean. Comparison of the impact structure and related deposits with those on land shows several major differences of which the most significant is the absence of an elevated crater rim. Instead, the crater perimeter is bevelled and eroded as a consequence of impact induced bottom currents and turbulent, return water flow into the excavating cavity. By this process most of the fall-out breccia is reworked back into the crater cavity where it accumulates in much larger thickness than in impact craters on land. At a microscopic scale, the shock metamorphism features are about the same as those for land impacts. Geochemically, impacts of comet nucleii, may not leave a recognizable signature at the impact horizon except for a minor increase in iridium. Thus stratigraphic horizons associated with extinctions and/or major changes in biota have to be closely examined for other impact indicators, like the presence of tectites, glass spherules, and quartz grains with shock features. Occurrence of megatsunami wave deposits, extensive erosion on continental margins, margin failures and faunal mixing above erosional unconformities are other potential impact indicators. There is no single indicator that can provide sufficient proof of an impact event. Such interpretations have to be based on multiparameter studies of global extent, since many of the impact indicators are only of regional extent. The lack of extinction of any marine plankton genera, or of bottom dwellers at the Montagnais impact site allows us to place a lower limit for biological extinctions caused by cometary impacts on those with nucleus >4 km in diameter. The calculated frequency for a cometary impact which could result in a 10% extinction of marine genera is about 6 × 10 −7 yr −1 and for the K/T boundary type of extinctions about 2 × 10 −8 yr −1. Even allowing for a large degree of uncertainty in these estimates, it is unlikely that the biological extinction events for the last 250 Ma identified by Sepkoski (1990) could have been all caused by meteorite impacts.