Research into swimming pool water treatment has been conducted on both bench and pilot scales. Initial work made use of a recirculatory bench-scale rig containing about 45 l of water. A more recent study was based on a 2.2 m 3-capacity pilot plant, a one-seventh linear scale model of an actual operational pool, incorporating conventional unit operations for swimming pool water treatment. The study is based on the use of a body fluid analogue (BFA), comprising the key dissolved organic and inorganic components of human perspiration and urine. Initial bench-scale trials indicated the analogue to be a reasonable representation of real body fluids in terms of levels of the key disinfection by-products (DBPs) of trihalomethanes (THMs) and chloramines (CAs) generated from conventional chemical disinfection by chlorination. Trials on the pilot-scale pool, on the other hand, indicated that: measured THM levels were generally much lower than those recorded on the bench scale, attaining equilibrium values after about 3 days of operation; chlorine demand, on the other hand, was about the same; for constant rate dosing of BFA the change in DBP levels on changing key operational determinants by ±50% was shown to be significant only on increasing the BFA dose rate (33% increase in THMs) and decreasing the disinfectant dose (50% decrease in THMs), the CA levels being largely unaffected throughout; dosing of BFA at a fluctuating rate (or intermittent) over a 14-h operating cycle produces higher levels of THMs than constant rate dosing at the same overall daily rate; the equilibrium ratio of chlorine arising in dissolved chloramines to that in dissolved THMs is around 42:1; a mass balance on the steady-state system can be carried out according to an assumption of either zero breakpointing or complete breakpointing to define the limits of DBP accumulation; the zero-breakpoint mass balance accounts for 38% of the organic carbon, 61–63% of the chlorine and 84–∼100% of the ammoniacal and amino nitrogen, the exact figures being very sensitive to assumptions made regarding trichloramine formation; the complete breakpoint analysis accounts for 52% of the chlorine for 100% oxidation of the ammoniacal and amino nitrogen, the organic carbon usage being unchanged at 38%. It is surmised from the analyses undertaken that, even at the 100% efficiency limits of Henry's law governed physical desorption or total breakpoint chlorination, a net accumulation of chlorine and carbon arises. More rigorous analysis of both the pool water and the ventilated air would reveal the extent to which such an accumulation actually takes place.