Satellite data are continuously used to monitor albedo and fraction of absorbed photosynthetically active radiation (FAPAR), which are key components in determining the energy balance and productivity of forests. However, due to the mismatch between spatial resolution of the satellite data and forest stand size, coarse resolution satellite products cannot capture the fine-scale variations in forest structure. Therefore, forest radiation budget models are important tools in quantifying albedo and FAPAR at stand scale. However, due to the lack of suitable input data, simulations are often restricted to summer conditions only and the seasonal patterns are not considered. We modeled the time series of albedo and FAPAR for an entire growing season for 20 forest plots in the boreal zone in Finland (61°50′ N, 24°17′ E) using an exceptional ground reference data set. Canopy gap fractions and the spectra of forest floor were monitored in the plots throughout the growing season. Data on the seasonality of spectra of tree foliage were also available. The modeled albedo and FAPAR were upscaled and compared against albedo and FAPAR derived from MODIS satellite data. We showed that forest radiation budget models capable of adequately taking into account foliage clumping and its effects on multiple scattering are the most appropriate for simulating albedo of boreal coniferous forests. Our results also indicated negative albedo-productivity relations in boreal coniferous forests. In addition, we demonstrated that not only the overall level, but also the seasonal patterns of albedo and FAPAR differ between tree species. Therefore, the use of only peak growing season albedo or FAPAR values when estimating climate impacts of forest management can be misleading.