During the aging process longitudinal changes for sustained work in humans are poorly understood. Only a few longitudinal follow-up studies have been published thus far, and most of them have been based on highly selected groups, such as physical education teachers or senior athletes. On the basis of random samples of a general population, the purpose of this study was to analyze changes in bicycle test performance in two longitudinal designs: a 10-year follow-up period from 50 to 60 years of age, using a submaximal test protocol, and another 10-year follow-up period from 70 to 80 years of age, using indirect voluntary maximal tests. In addition, the preventive value of the bicycle test results for survival at different age levels was analyzed. The subjects in the first part of the study were members of a random sample of originally 514 men and 461 women living in the Glostrup area, close to Copenhagen, Denmark, in 1964. Of these, 367 men (71.4%) and 206 women (44.7%) were tested at the age of 50 years, and 309 men and 245 women were tested 10 years later. The subjects in the second part of the study came from the same original sample. At 70 years of age 171 men and 154 women and at 80 years of age 70 men and 68 women took part in the maximal test. The submaximal test results between the ages of 50 and 60 years showed a mean annual decline in body mass adjusted maximal power in sustained work (W/kg) of 0.54% in men and of 0.90% in women. Between the ages of 70 and 80 years, when the indirect maximal tests were applied, the annual decline in men was on average 1.79% and in women 3.03%. When the associations of submaximal test results at ages 50-60 years and the voluntary maximal test results at the higher ages were analyzed, a moderate positive correlation was observed with the results obtained at the age of 70 years. The survival analyses showed that the submaximal bicycle test results (W/kg body mass) at the age of 60 years had a predictive value for survival in women during the subsequent 10-year period. The same was true for the maximal test results obtained at the age of 70 years in men; a significantly larger proportion of men in the lowest quintile died during the subsequent 10 years than of those belonging to the higher quintiles. The changes in body mass related maximal power in sustained work observed in this population on the basis of longitudinal studies among the age groups 50-60 and 70-80 years indicated a steeper decline at the higher ages. The decline was relatively more pronounced in women than in men. However, differences in the test protocols employed at different times limit the possibilities for overall comparisons across the data. The results of the submaximal bicycle ergometer tests in middle-aged female subjects (60 years old) had a predictive value for survival over the 10 years immediately following the test; likewise, the voluntary maximal test results at higher ages predicted survival in men.