The radar meteoroid orbit data set obtained from the AMOR facility in Christchurch, New Zealand (longitude 172°39′ E, latitude 43°34′ S) between 1995 May and 1999 October contains ∼5 × 105 high-quality meteor records. The system was very sensitive compared with previous surveys, with a limiting radio magnitude of +14 corresponding to a 3 × 10−10 kg meteoroid mass limit (40-μm diameter) being achieved. This data set is here examined to determine and remove biases inherent in the radar method. The fully corrected meteoroid orbital distribution at 1 au from the Sun is derived. This distribution replaces a previous, much used, orbital distribution produced by the earlier Harvard Radio Meteor Program (HRMP). Anomalies have been found in the original debiasing of the latter which strongly favoured meteoroids observed at low speeds. Three forms of output orbital element distributions have been produced in the present study. To aid comparisons, these forms are identical to those produced by the HRMP: the ‘directly observed’ output is that with no corrections applied, the ‘atmospheric’ sample is corrected for all in-atmosphere effects, which includes electromagnetic wave propagation and ionospheric effects and the influence of the particular form of the radar system, and the ‘space sample’ is additionally corrected for collision probability with the Earth. The space sample has rather higher eccentricity and larger semimajor axis length orbits than directly observed by AMOR. Its inclination distribution shows a general decrease in number with inclination: with a peak at ∼20° and few meteors at inclinations very close to the ecliptic; a small population remains from the original ∼50 per cent of orbits in retrograde orientations. Comparison with the original HRMP space distributions shows little agreement, however the revised HRMP orbital element distributions of Taylor and Elford compares well. The higher number of orbits in the AMOR data set and the uncertainty involved in re-reducing the older HRMP data shows a strong direction for the use of the AMOR as a standard.