Three real-time, instream water-quality and turbidity-monitoring sites were established in October 1998 in the upper North Santiam River Basin on the North Santiam River, the Breitenbush River, and Blowout Creek, the main tributary inputs to Detroit Lake, a large, controlled reservoir that extends from river mile 61 to 70. Suspended-sediment samples were collected biweekly to monthly at each station. Rating curves provided estimated suspended-sediment concentration in 30-minute increments from log transformations of the instream turbidity monitoring data. Turbidity was found to be a better surrogate than discharge for estimating suspended-sediment concentration. Daily and annual mean suspended-sediment loads were estimated using the estimated suspended-sediment concentrations and corresponding streamflow data. A laboratory method for estimating persistent (residual) turbidity from separate turbidity samples was developed. Turbidity was measured over time for each sample. Turbidity decay curves were derived as the suspended sediment settled. Each curve was used to estimate a turbidity value for a given settling time. Medium to fine clay particle (< 0.002 mm [millimeter] diameter) settling times of 8.5 hours were computed using Stokes Law. An average of 30 persistent turbidity samples was collected from each of the 3 sites. These samples were used to estimate the 0.002-mm-size clay particle persistent turbidity for each site. The monitored instream 30-minute turbidity values were converted to a calculated persistent turbidity value that would have resulted after 8.5 hours of settling in the laboratory. Persistent turbidities of 10 NTU and above were tabulated for each site. (Water of 10 NTU and above can interfere with or damage treatment filters and result in intake closures at drinking-water facilities.) A method was developed that used the persistent turbidity experiments, turbidity decay curves, and stream discharge to estimate the volume of water containing suspended clay that entered Detroit Lake from the three main tributaries. 'Suspended-clay water' was defined as water having a value of at least 10 NTU after settling the required 8.5 hours. The suspended-clay concentrations of 10 NTU or higher were paired with the corresponding stream discharge in the continuous record. These summed discharges represent the annual volume of water containing suspended clay that entered Detroit Lake from the three main tributaries. Higher yields (load per unit area) of suspended sediment and suspended-clay water were observed from the smaller Breitenbush River and Blowout Creek subbasins than from the main-stem North Santiam River for water years 1999 and 2000. The 3-day peak streamflow and turbidity events in 1999 and 2000 carried two-thirds of the annual suspended-sediment load for the three subbasins. Turbidity and suspended-sediment concentration relations within the upper North Santiam River Basin are basin specific and can change annually within a single subbasin. Techniques developed during this study will assist water resource planners in understanding and managing water quality in their watersheds, particularly those in which there are persistent-turbidity problems.