Physiological saline (0.9%) was administered intramuscularly into the longissimus dorsi muscle of rabbits (four animals per group) at different injection speeds (1 ml/sec. or 1 m1/10 sec.) and at different dose volumes (2 or 6 ml). The injection sites were examined post mortem 30 minutes after the injection. The same procedure was used for the administration of a solution of 1 % cis(Z)-clopenthixol, a neuroleptic drug, in 0.9% NaCl at speeds of 2 m1/2 sec., 2 m1/6 sec., 2 m1/20 sec., or 2 m1/60 sec. The injection sites of these rabbits were examined post martem 3 days after the injection. The muscle tissue at the injection sites was examined for macroscopic changes by serial sectioning. If a demarcated area of necrotic tissue was found, the necrotic area was dissected from the surrounding muscle tissue and weighed. Muscle tissue from rabbits receiving an injection of 0.9% NaCl was prepared for microscopic examination. In almost all rabbits given 0.9% NaCl the needle track was indicated by slight bleeding. In some of the rabbits which received 2 o r 6 ml of 0.9% NaCl over 2 or 6 sec., slight bleeding was also seen in the muscle tissue adjacent to the needle track. Since no further changes were seen, it is concluded that very fast injection of 0.9% NaCl only causes slight bleedingprobably fromdisruption of the local tissue. The needle track was identified microscopically by haemorrhages and disruption of muscle fibres. Muscle bundles were separated and muscles fibres were eosinophilic or vacuolated. In the rabbits given 1 % cis(Z)-clopenthixol, a well-demarcated area of necrotic muscle tissue was found at the injection sites. There was no difference between the mean weights of necrotic tissues from the groups given 2 ml over 2.20 or 60 sec. (mean weightsfS.E.M., 5.15f0.41, 4.6750.03, or 4.59f0.15 g, respectively). However, the mean weight of necrotic tissue from animals given 2 ml over 6 sec. (5.69f0.24 g) was significantly higher than that of the mean weights after injections given over 20 or 6Osec. From these findings it is concluded that injection speed is not an important factor for local damaging effect of intramuscular drug preparations. However, very slow injection may reduce the local damage but only slightly.