Perennial organ functions of trees living in seasonal environments exhibit temporal changes that can be classified as long-term interannual changes and seasonal fluctuations within single years. However, few studies have separately quantified these changes from longitudinal measurement data or analyzed the relationships between them. We developed a hierarchical Bayesian statistical model consisting of three parts: a long-term interannual change expressed by consecutive annual linear trends, seasonal fluctuations with 26 values for two-week periods in a year, and a random effect for repeated measurements. The model can extract long-term interannual changes and seasonal fluctuations from longitudinal repeated measure data. The pattern of seasonal fluctuation, the amount of seasonal fluctuation, and the net annual change are expressed by the estimated model parameters. We applied our model to foliar chlorophyll (Chl) and nitrogen (N) content measured repeatedly on more than 1-year-old leaves of saplings in four evergreen broad-leaved tree species using nondestructive optical methods. The model successfully explained large variations in the Chl and N content. In general, seasonal fluctuations corresponded to the phenology of current-year leaves; Chl and N tended to decrease from the opening to maturation of new leaves and increased during the rest period. The magnitude of the decrease in the Chl and N content in the growth period of current-year leaves (Δγ) did not decrease noticeably as leaves aged. For the Chl content, Δγ was positively correlated with the maximum value before leaf opening across species. For the N content, Δγ and the maximum value before leaf opening were not clearly correlated across species, but were positively correlated within some species. A model parameter for annual linear trends in Chl and N varied from positive (indicating increasing trends) to negative values (indicating decrease) depending on species and leaf age in years.