This paper presents a demographic study of mortality levels, differentials, and trends, using mortality data collected in the 1977 Lesotho Fertility Survey (LFS), and 3 earlier surveys. The LFS consisted of 1) a household survey of 86,744 people, 2) a more detailed subsample of 3603 ever-married women, and 3) a post-enumeration survey to test response reliability. The LFS estimates infant mortality as somewhat over 110/1000 births. Death probability before age 5 declined from 20% in the mid-1960s to about 17.5% 10 years later, and data suggest a slow but steady decline beginning in the mid-1950s. The high infant mortality and low child mortality patterns are somewhat atypical for sub-Saharan Africa; at least 50% of infant deaths occur in the 1st month of life. LFS data show seasonal variation in infant mortality; 18.5% of children born in the autumn and winter die before age 5, compared to 16% of children born in the spring or summer. Survey results include such infant and child mortality differentials as 1) boys are 10% more likely to die before their 1st birthday than girls, 2) 3-4% more children between birth and age 5 die in eastern Lesotho than in the Lowlands or Foothills, 3) the children of women with no schooling are more than twice as likely to die between ages 2 and 5 as the children of women with a secondary school education, 4) 18% of children with uneducated fathers die between birth and age 5 compared to 14% of children with fathers who have a secondary school education, 5) women in their 20s are less likely to lose children than those in their teens and 30s, and 6) birth intervals of less than 2 years increase mortality for the earlier born as well as the subsequent child. Lesotho's 4 demographic surveys also provide direct and indirect estimates of adult mortality; they suggest that by the mid-1970s, life expectancy at age 15 was 45 years for males and 55 years for females. Little reduction in adult mortality was experienced between the 1950s and the 1970s. Differentials in adult mortality include the fact that 1) adult mortality varies little by region, and 2) increased education increases life expectancy.