For over 100 years, synthetic chemicals have been added IO foods solely to enhance theii physical appearance. Only recently have rational rixk-benefit assessments been coGdered for foad additives. In the mid-1970s Feingold, a clinical allergist, contended that almost one-half of children diagnosed as having chiklhond hype&k&s and leaming disorder (mini. mal brain dysfunction, attention-deficit disorder) showed improvement when placed on adiet that eliminated all artificial colas and tlavors. some anifcial pieserva tives, iztd ail foods ~~n~g a ‘natural salicylate’ mdicaP. Feingold’s original elimination diet was based on clinical case descriptions of i&It patients who experienced allergi4ike symptoms in response to aspirin, When the subjects’ co~ions faifed to impmve after the elimination of exposure’to aspirin-like compounds, they were placed on a salicylatefre~ diet that prohibited the ingestion of ail saiicylates. ~~l~i~ tboe thought to OcEur naturally in fcrads. The artif&l fbod color tar5zzine (FIMC Yellow No. 5) was alsoeliminated from the diet because it produced the same &z&c-like symptoms as salieylates in ~~~sidve patients. Subsequently. the diit was expanded to exclude ail food additives, including those chemically unrelated to aspirin. Anecdotal rem* sumted that this diet not only ivies ~~e~~like ~tivi~ to aspirin in susceptible patients. but also impmved the behavioural symptoms of psychologi. caIiy distutbed patients. Feingold then began using the diet to treat children with bebavioural disturbances, st~~e~g that somfi causal relationship miust exist be. hueen the increasing amount5 of artificial additives u5ed in foods over the last several decades and the imzasiig number of ciinlcal reports of chifdhrnxl hyperkinesis 81x1 learning disorder over ti same time poiod. The original Feingold hypothesis and diet have undergoM sub5eqlIGrt revisions. The mod&d d&t now exchtdes just anifici colors and flavors. although mos$ Peingold diets studied in clinical tests omit the prohibition of flavors owing to difficulties with disguising the flavor3 in placebo diets and cha&nges, and with i~l~ing the suspected offending compound from the tiusands of flavoring constituents in use. The major questions to be re.solved. -fore. aze whether f& sddiives do cause obvious changee:_=_I whether any efiect of such additives can be demonstrated on r~urocnrhemical systems. Feingold claims that many cases of childhood hyperkinesis and learning disorder can be explained as ~ios~tic, not allergic. reactions IO food colors which manifest themselves as behavioural responses in genetically susceptible chiidrcn. Similar sensitivities to f& have been known to occur. T?K symptotns of f&d ~nsiti~~ ate often vague and variable, including aches and pains, fatigue and behavlour&l disturbances. Childhood hyperkinesis and leaming dis~rc~~~st~~~~aci~t~of symptoms. Children diagnosed as hyperkinetic usually show elevated levels of motor activity, poor control ot impulsive behaviour, restlessness, a shwt attention span, a low tolerance for fanatic, irritability. excitability and aggressiveness. These symptoms are not, however, exhii bited uniformly and it is difficult to determine whether chilw hyperkinesis is a discrete illness or a synch with several biochemical or physiological bases. Treat-