Yes, we do this all the time. The method was developed by George Wilcox. We used it in these papers:
Walwyn, W, W Chen, H Kim, A Minasyan, H Ennes, J A McRoberts, J C Marvizon. Sustained suppression of hyperalgesia during latent sensitization by µ, δ and κ opioid receptors and α2A adrenergic receptors - role of constitutive activity. J Neurosci 36: 204-221 (2016). PMC4701961
Severino, A, W Chen, J K Hakimian, B L Kieffer, C Gaveriaux-Ruff, W Walwyn, J C G Marvizon. Mu-opioid receptors in nociceptive afferents produce a sustained suppression of hyperalgesia in chronic pain. Pain 159: 1607-1620 (2018). PMC6053329
The right volume for an intrathecal injection in mice is 5 ul, with a 5 ul chaser. Otherwise, you risk the injectate spilling over into the brainstem.
The procedure is not easy because you have to get a feel of when the tip of the syringe is under the dura but still out of the spinal cord. The angle of the needle is crucial. The way my investigators train for this is by injecting morphine. If the injection is correct, the mouse will develop a Straub tail: the tail is erected vertically.
We do it without anesthesia. One person holds the mouse inside a cloth while the other does the injection.