Over a period of 12 years Denhoff and Holden (1961) tested 8 drugs on pre-school children suffering from cerebral palsy for improvement in neuromuscular function including spasticity, behaviour, taste (acceptability) and toxicity. Objective tests were used for the first. The drugs examined were mephenesin (Myanesin, Tolserol), mephenesin with glycerol-alpha-ethyl-gamma-isopropyl ether (Lissephen), chlorpromazine, reserpine, zoxazolamine (Flexin), alpha-ethyl-beta-methylvaleramine (McN 181), carisoprodol (Soma), and emylcamate (Striatan). Beneficial results were seen in neuromotor function with both mephanesin and placebo. Placebo effects were greater than drug effects in neuromotor function and behaviour with Lissephen, but this drug was not tolerated in doses sufficient to reverse these effects. With chlorpromazine behaviour improved statistically, but the neuromotor effectiveness was secondary, occurring because the child was mentally relaxed and happy. Zoxazolamine gave improved behaviour, but the neuromotor effect was no better than that of the placebo. Reserpine was not tolerated, drowsiness taking place with a dose necessary to produce efficient performance. McN 181 was ineffective in doses tolerated and 6 cases were unchanged or worse as regards spasticity on soma. Striatan gave 6 of 14 improving in anxiety, but 10 of 18 did so with placebo and no statistical differences were found in anxiety score. The beneficial neuromotor effects of the drugs varied from 9 to 56 per cent. (mean 33.4) and of the placebos from 26 to 50 per cent. (mean 38.7). The authors point out that overall progress can be made when the child is mentally relaxed and happy, as when being treated with chlorpromazine.