Action-potential characteristics and muscle force were investigated in forty-five patients in repeated examinations during the initial stage of poliomyelitis and one year after the onset of the disease. The symmetrical normal muscles and 130 normal subjects served as controls.

In order to evaluate the prognostic value of action-potential analysis, the patients were grouped according to the degree of initial paresis and recovery from it one year l ater, as obtained by muscle grading.

In the acute stage of the disease, the mean action-potential duration was increased 30 per cent. in severely affected muscles which did not show signs of recovery one year later (10.5 ± 0.4 milliseconds as compared with 8.2 ± 0.2 milliseconds in normal muscles). Muscles which were equally severely paretic in the acute stage, but which improved greatly in the course of one year, had normal values of mean action-potential duration. Muscles with a less severe paresis but with no or slight recovery had a normal mean action-potential duration. One year after the acute stage, all affected muscles, independent of the degree of recovery, showed an increase in mean duration of approximately 20 per cent. above normal.

There was a considerable increase in the incidence of polyphasic potentials in the acute stage of the disease. This increase could not, however, account for the increase in mean action-potential duration.

The mean action-potential amplitude recorded during the initial stage of the disease was within normal limits (200 microvolts) for all groups and was increased by 50 to 150 per cent. one year after the onset without correlation to the degree of recovery.

Fibrillation potentials (spontaneous low-voltage activity of short duration) occurred in 62 per cent. of all paretic muscles in the initial stage and in 35 per cent. one year after the onset of the disease. The earliest sign of denervation was found thirteen days after the onset of the disease. In acutely paretic muscles which later recovered slightly or not at all, fibrillation potentials occurred more frequently (75 per cent.) than in severely paretic muscles which showed good recovery (33 per cent.).

Synchronous activity between different leads in the same muscle was clearly correlated to the degree of recovery. It occurred in 40 per cent. of paired leads in severely paretic muscles which did not recover, examined in the acute stage, as compared with 20 per cent. in the other groups and in normal controls.

In the initial stage of the disease, the thermo-electrically measured intramuscular temperature was significantly decreased (0.9 degrees centigrade) in severely affected muscles without recovery, as compared with the other groups and normal controls. One year later there was a similar tendency.

The finding of a correlation between muscle action-potential characteristics on the one hand and degree of recovery on the other indicates that the eventual outcome of the muscle paresis is largely determined by the initial disease process. Early electromyography may thus provide a guide as to which muscles will and which will not benefit by physical therapy directed to those muscles.