A full analysis of radar-determined meteor rates from New Zealand, involving 3,085,574 meteors recorded over a total of 3 1/2 years, and 12,391,976 meteors recorded by the National Research Council of Canada in 8 1/2 years confirms an inverse relationship between meteor rates and solar activity as measured by sunspot numbers. The relationship, significant at the 1% level, appears in the Canadian annual average when the abnormal 1963 increase is removed, in monthly and 1/3-monthly results for the total Canadian period, and in monthly intervals for 1 year of the New Zealand data. This proven relationship of meteor rates with the solar cycle calls for a significant density gradient change over the solar cycle in the 70- to-120-km height range. Although some definite negative results have been reported, no unambiguous positive results are yet available supporting such a density gradient change. It is possible that density variations due to annual, semiannual, diurnal, and latitudinal changes obscure any 11-year density gradient change occurring at these heights. It is uncertain whether the 1963 increase represents density gradient changes in the meteor ablation region regularly brought about 1-2 years before each sunspot minimum or is a special event due to volcanic dust. The following additional facts have emerged from the present analysis. (1) Within a 1-year period the seasonal rate change of astronomical origin overrides any density gradient change in controlling the meteor rates in one of the two hemispheres. (2) The earth's daily rotation alters rates in phase with probable diurnal density gradient changes. (3) An effect due to D region absorption has been observed in the Canadian data. (Such absorption can alter radar meteor rates appreciably in the lower tens of megahertz.)