Ityoscine and also arsenic in large doses were long considered to be effective in HV~TI~GTO~'S Chorea (SI~xLwR 1889) and similarly hyoscyamine, atropine and scopolamine combined in a proprietary preparation (~abellon) were favourably described in recent years by To~aLI~so~ (1947). Interest in possible drug therapies for this condition appears to have increased however since the paper of GOLDMAN (1952) in which the use of procaine amide was recommended, although subsequent investigations, reviewed by M]~RSK]~Y (1958) gave only limited support to GOLDMAN'S claim. Reserpine, which was first used by CHA~D-LV, R (1955), has been accepted as effective by a number of authors (LAzARTE et al. 1955; BRAND~ 1956; K]~PI~SKY et al. 1960) who often used cinematographic techniques to verify their conclusions. But a high incidence of undesirable side-effects has typically been found in such studies with reserpine. For current treatment the choice appeared to I)AKE~]~ABI-WA]LSH (1960) to lie between thiopropazate recommended by MATT~WS (1958) and perphenazine, described for one case by LA~ICCHIA and BAS~I (1958). In addition the thorough study of M:~TE~s and F~SCH~lz (1958) showed clear clinical benefit in choreic states from the use of both intravenous and oral perphenazine. We therefore decided to attempt an assessment of the effects of perphenazine in four cases. Methods The trial was originally designed according to a double-blind pattern, the key to the placebo and drug being kept by the Hospital Pharmacist. In the event the emergence of side-effects with substance "A" made us confident that the trial was no longer "blind" and we therefore confined our observations to the state o~ the patients before, during and after the oral administration of this drug which we believed, correctly as it turned out, to be perphenazine. The dose of the drug was varied in the individual cases so as to obtain the maximum level which appeared beneficial without causing side-effects. This ranged from 8 mgms t.d.s.