A series of jar tests using iron (III) chloride were carried out to identify factors affecting phosphorus removal. Mixing conditions and the extent of iron hydrolysis prior to reaction with phosphate ions were explored using crude sewage as a medium. Somewhat surprisingly, the mixing conditions appeared to have little affect on the phosphorus removal performance. As expected, there was a strong link between phosphorus removal and iron:phosphate ratio during standard jar tests with iron (III) chloride. Total phosphorus removal of 80% was achieved at a dose of 1.48:1 molar ratio Fe:P. When a mixture of pre-hydrolysed iron and iron chloride was added, 80% total phosphorus removal was achievable at a dose of 1.86:1 molar ratio Fe:P. These findings could make a considerable difference to the cost of iron (III) chloride addition for phosphorus removal, if the hydrolysed product is formed preferentially to the phosphate. Increased costs could amount to more than pound60,000 for a site treating a population equivalent of 200,000 and requiring a 1 mg l(-1) total phosphorus effluent target. In reality, the hydrolysed product is unlikely to form to this extent but the results suggest that the formation of hydroxide may not only take place when the phosphorus removal reactions are complete as the relative reaction rates are more similar than perhaps was previously thought. Metal hydroxides have the capability to remove phosphorus by absorption but less efficiently.