An examination of past human mortality trends revealed that the mean longevity of birth cohorts from 1740 to 1900 for United States of America (U.S.) Congressional Representatives exhibited oscillations that coincided with the 9- to 12-year sunspot cycle. Cohort mean longevities were 2-3 years greater during times of low sunspot activity than at peak activity. This phenomenon was confirmed in data from members of the House of Commons of the United Kingdom Parliament and from University of Cambridge alumni. An additional longevity oscillation with a longer period was visible in the data and may also be related to sunspot cycles. The amplitude and frequency modulations in the longevity and sunspot oscillations aligned when a 20-year phase shift was incorporated. This shift requires the existence of a lag between solar changes and the affected birth cohorts. Several possible causes of the effect are discussed, in particular: radiation on primordial germ cells in developing embryos; influenza epidemics and pandemics; and weather. The size of the longevity oscillation requires that the solar effect must be considered in studies that examine longevity trends and risk estimation.