STUDY OBJECTIVE--The aims were (1) to monitor and compare incidence rates of cancer from successive birth cohorts in Saarland over the period from 1968 to 1987; (2) to project cancer incidence in Saarland in 1988-2002 in order to provide guidelines for health policy planning. DESIGN--This was an ecological study of overlapping birth cohorts of women and men. SETTING--The study was population based involving the whole state of Saarland. PATIENTS--80,028 cases of malignant neoplasms (other than non-melanoma skin cancer) diagnosed from 1968 to 1987 and reported to the cancer registry of Saarland were included. MEASUREMENTS AND MAIN RESULTS--Age specific, sex specific, and period specific cancer incidence rates were analysed and extrapolated by multiplicative age-cohort models. Due to a steady rise in birth cohort specific cancer incidence rates in males, a substantial rise in incidence of total cancer is projected, while a moderate decline is expected for females. Analogous analyses are presented for the most common single forms of cancer in women and men. Alternative strategies of analysis, such as age-period-cohort modelling, are discussed. CONCLUSIONS--The age-cohort model is well suited for monitoring incidence of most forms of cancer. The projections provide quantitative guidelines for planning of health care resources and underline and quantify the challenge for primary and secondary cancer prevention in Saarland.