The bulk of meteoroidal particles follow pseudo-random orbits and are termed sporadic meteoroids. These are thought to be derived from the correlated streams of particles released by comets, although the mechanisms by which their orbits are dispersed have been the subject of some confusion.

By developing techniques to compute the frequency of close encounters with each of the planets, and also the gross outcome of such events, it is shown that most sporadic orbits are a result of gravitational scattering by the giant planets. Jupiter plays the major role.

Although catastrophic impacts with smaller particles limit the lifetimes of meteoroids, this mechanism is not responsible for the bulk of the stream disruption. With a simple model of the zodiacal cloud, the method is also used to find the collisional lifetime of meteoroids including for the first time the dependence upon inclination. The rate of meteoroid depletion by planetary collisions and hyperbolic ejections resulting from close approaches is calculated. It is found that for Jupiter-crossing meteoroids these losses are as rapid as those due to the PoyntingRobertson effect.

This theory is also applied to six peculiar asteroids, including Hidalgo and Chiron. These prove to have extremely short-lived orbits: large orbital variations occur on a timescale of only ~10³ years. It is also shown that Pluto exists in its Neptune-crossing orbit solely because of the stable resonance which prohibits approaches between the two in the present epoch.

The collision rate between the Apollo-Amor-Aten asteroids and each of the terrestrial planets is calculated using all 76 known objects. The result using this new procedure (4-6 Earth impacts per million years) is somewhat higher than previous estimates, indicating that these asteroids do not represent a steady-state population.