We describe a new test system for exposing fish to resuspended sediments and associated contaminants. Test sediments were resuspended by revolving test chambers on rotating shafts driven by an electric motor. The timing, speed, and duration of test-chamber revolution were controlled by a rheostat and electronic timer. Each chamber held 45 litres of water and accommodated about 49 g of test fish. The system described had three water baths, each holding six test chambers. We illustrate the performance of this system with results from a 28-day test in which juvenile bluegills Lepomis macrochirus were exposed to resuspended, riverine sediments differing in texture and cadmium content. The test had one sediment-free control and five sediment treatments, with three replicates (chambers) per treatment and 25 fish per replicate. Two-thirds (30 litres) of the test water and sediment in each chamber was renewed weekly. The mean concentration of total suspended solids (TSS) did not vary among treatments; the grand-mean TSS in the five sediment treatments was 975 mg litre⁻¹, similar to the target TSS of 1000 mg litre⁻¹. At the end of the test, an average of 50% of the introduced cadmium was associated with the suspended sediment compartment, whereas the filtered (0.45 μm) water contained 0.4% and bluegills 1.8% of the cadmium.