Turbidity is commonly used as a proxy to estimate suspended sediment content in streams, and for hydroecological purposes. Uncalibrated turbidity records in conjunction with water chemistry data prove useful in detecting watershed reaction to single hydrological events during the spring flood in Arctic Russia. The turbidimetric survey technique was applied to study the spatial variability of sediment yield features on small rivers of the south-eastern part of Sakhalin Island. Suspended sediment concentration (SSC) vs. turbidity relation follows the geological features of the terrain and reflects the land-use intensity within the watersheds. For our Igarka key site, a logarithmic regression model was developed as an instrument of SSC calculation with turbidity data for each of the four studied watersheds. Our results show that factors influencing turbidity, namely water color and sediment grain size, have to be considered in multivariate models, to minimize errors and acquire an understanding of what kind of physical response is actually measured by nephelometry-based instruments. This study was held within SEDIBUD work group in 2011-2013.