This paper describes the occurrence and levels of gingival recession in 2 cohorts of individuals participating in parallel longitudinal studies in Norway (1969-1988) and Sri Lanka (1970-1990), covering the age range from 15 to 50 years. In the Norwegian cohort gingival recession had begun early in life. It occurred in greater than or equal to 60% of the 20 year-olds and was confined to the buccal surfaces. At 30, greater than or equal to 70% had recession, which still was found mainly on buccal surfaces. As the group approached 50 years of age, more than 90% had gingival recession; greater than or equal to 25% of the buccal surfaces were involved, greater than or equal to 15% of lingual, and 3 to 4% of the interproximal surfaces. In the Sri Lankan cohort greater than or equal to 30% exhibited gingival recession before the age of 20 years. By 30 years, 90% had recession on buccal, lingual, and interproximal surfaces; and at 40 years, 100% of the Sri Lankans had recession. As they approached 50 years, gingival recession occurred in greater than or equal to 70% of the buccal, greater than or equal to 50% of the lingual, and 40% of the interproximal surfaces. Based on the special features of the two cohorts, the working hypothesis is advanced that there is more than one type of gingival recession and probably several factors determining the initiation and development of these lesions.