1. The synthesis of acetylcholine (ACh) has been measured in homogenates of the sciatic nerve, normal and denervated extensor digitorum longus (e.d.l.) muscles, and central (innervated) and peripheral (non-innervated) parts of the diaphragm of the rat. The synthesis proceeded under conditions accepted as optimal for the activity of choline acetyltransferase (ChAT). In view of the finding that cardiac carnitine acetyltransferase (CarAT) is able to acetylate choline (White & Wu, 1973), the possible contribution of CarAT to the synthesis of ACh in the muscles was investigated by using bromoacetylcholine (BrACh) as an inhibitor of ChAT and bromoacetylcarnitine (BrACar) as an inhibitor of CarAT.2. BrACh at a concentration of 2 mum inhibited the synthesis of ACh in nerve homogenates by 98%, in the homogenates of normal e.d.l. muscles by 53%, in denervated e.d.l. muscles by less than 5%; in the central part of the diaphragm BrACh inhibited ACh synthesis by 65%, and in the peripheral part by 13%. Comparative inefficiency of BrACh in inhibiting the synthesis of ACh in muscle homogenates was not due to its inactivation; the inhibitory effect of BrACh on the neural synthesis of ACh was preserved in the presence of muscle homogenates.3. BrACar at a concentration of 20 mum inhibited the synthesis of ACh in homogenates of the nerve by 18%, in those of normal e.d.l. muscles by 67%, in denervated e.d.l. muscles by 90%; in the central part of the diaphragm it inhibited the synthesis by 29%, and in the peripheral part by 76%.4. The inhibitory effects of BrACh and BrACar on the synthesis of ACh in muscle homogenates were roughly additive.5. Within 2 days of transection of the sciatic nerve, the BrACh-sensitive synthesis of ACh in the e.d.l. muscle diminished by 28%, whereas the BrACh-insensitive synthesis of ACh did not change. At 4 days after denervation, the rate of BrACh-sensitive synthesis decreased to 3% of control values.6. The results indicate that at least two enzymes are responsible for the synthesis of ACh in muscle homogenates is probably catalysed by CarAT. associated with intramuscular nerves, and probably corresponds to ChAT. The other enzyme is comparatively insensitive to BrACh, is sensitive to BrACar, and is probably localized in the muscle fibres. The BrACh-insensitive and BrACar-sensitive synthesis of ACh in muscle homogenates is probably catalysed by CarAT.7. Under the conditions used in the present experiments, CarAT was responsible for approximately one half of the synthesis of ACh in the homogenates of innervated e.d.l. muscles and for all ACh synthesized after denervation.8. The results provide an explanation for earlier findings of residual ACh synthesis in homogenates of denervated muscles, without resort to the idea that ChAT is localized in muscle fibres. It is proposed that CarAT catalyses some synthesis of ACh also in intact muscles, and that it is responsible for the synthesis of ACh observed during incubation of whole denervated muscles. It is not clear what is the physiological function of the synthesis of ACh catalysed by CarAT.9. Measurements of the BrACh-sensitive portion of the total ACh-synthesizing capacity of muscle homogenates provide a suitable procedure for obtaining information about the activity of neural ChAT in the muscles.