Spectroscopic measurements, which are non-destructive, precise and rapid, can be used to predict soil properties and help to estimate the spatial variability of soil properties at different scales. We collected laboratory-based visible-near infrared (vis-NIR) spectra of permafrost-affected soils with various active layer depths at the pedon scale. To estimate the accuracy of measured soil properties using vis-NIR spectroscopy, we first measured chemical and physical soil properties of the samples (n=470). Nested samples (n=296) were possible to analyze only using vis-NIR measurements due to a small amount of soil material. Thus, vis-NIR spectroscopy allowed us to collect more information within the profile in order to map the spatial patterns using a geostatistical method. We can conclude that vis-NIR spectroscopy can be used effectively in order to describe, estimate and map spatial distribution and variability of soil properties. This research could also help to improve the global soil spectral library since few applications of vis-NIR spectroscopy were done in northern Siberian soils.