Results from the Brimstone and other drainage experiments have shown that drainwater and surface layer flow from clay soils under arable cultivation usually contain >0.02 mg phosphorus/l in both soluble and particulate forms. Addition of phosphorus fertiliser to wet soil can increase the concentration of soluble phosphorus in drainflow; however, reducing the amount of fertiliser applied to less than that required to compensate for uptake by the previous crop usually decreases the concentration of soluble phosphorus in drainflow only slightly. Treatments which have been successful in reducing soluble and total phosphorus losses at Brimstone are those in which phosphorus is absorbed from the water flowing through mole drains, either by increasing the contact time of water and subsoil or by use of active absorbent materials inserted into drains. Results have, however, been variable and further work is required on the effectiveness of these and other treatments under a range of winter rainfall conditions. In this study, effects of the following treatments on losses of phosphorus in soluble (molybdate-reactive) and particulate forms in drainwater and surface layer runoff will be measured in the hydrologically sealed plots of the Brimstone Experiment: 1. Two rates of phosphorus fertiliser application, one compensating for the previous crop’s uptake, the other at half this rate. Effects on soluble and total phosphorus losses, nitrogen loss, and crop growth and yield will be measured under different winter rainfall conditions; 2. Drainflow either unrestricted or restricted by inverted rotatable U-bends in the pipes carrying drainwater away from the plots; 3. Two different mole channel spacings (2 and 4 m); 4. With and without secondary autumn cultivation of the seedbed to create a finer tilth; and 5. Use of phosphorus absorbent materials (ironstone gravel, the natural organic material Jimsorb, an aluminium oxide-rich waste product of aluminium reclamation and barley straw) inserted into the mole channels or in pipes carrying drainwater away from the plots. Where the aluminium oxide-rich waste product is used, the soluble aluminium content of drainwater will be monitored. Results will be used to evaluate and improve the recently developed phosphorus -leaching model PSALM for predicting phosphorus losses from cracking clay soils, taking weather patterns into account. Since treatments which are successful in reducing the amount of one surface water contaminant may increase another, results will be integrated with those from other studies on pesticide leaching, the ultimate aim being to develop techniques that will deal effectively with all contaminants of agricultural origin.