|Excessive delivery of fine-grained sediment and sediment-bound nutrients to surface waters results in water quality impairment. Information on the relative contribution of different sources contributing sediment to river systems is a prerequisite to target management practices. Sediment fingerprinting technique can help to estimate sediment contributions from various sediment sources to fluvial sediment load. The overall goal of this study was to determine the sources of in-stream sediment (stream bed and suspended) at a subwatershed scale using sediment fingerprinting approach in an urbanized, 31 km2 Moore’s Mill Creek watershed in Southern Piedmont region in Alabama. The relative source contribution from construction sites and stream banks to in-stream sediment was quantified for two different particle size fractions, 63- 212μm (fine sand) and <63μm (silt and clay). Results of this study showed that both construction sites and stream banks were important sources of stream bed sediment. The stream bed sediment in the upstream reaches originated largely from channel bank sources, and in the lower reach (watershed outlet), construction sites were the dominant sources of stream bed sediment. Also, this study showed that the construction sites were the dominant sources of suspended sediment in the watershed with contribution ranging from 0 to 100%, varying temporally. The relative source contribution from different sources is dependent on the particle size of the sediment, time and location of sampling within a watershed, riparian buffers, and areas of construction activities in proximity to the sampling sites. Also, it was observed that different source contributions could be obtained with different fingerprinting procedures as apportionments are sensitive to the statistical procedures employed. Soil and Water Assessment Tool (SWAT) was used in parallel to assess valuable information of watershed-level hydrological processes that affect sediment erosion and transport within a watershed. SWAT identified areas that generate high surface runoff and water yield and have the potential to contribute disproportionately high amount of sediment to streams. Targeting best management practices (BMPs) in these areas can significantly reduce the sediment loadings to the streams. Overall, this study underscores the importance of considering the spatial and temporal variability of sediment sources as a function of sediment particle size for targeting BMPs. The combined use of sediment fingerprinting technique and watershed-level modeling can provide valuable information of sediment transport processes and dynamics within a watershed.