OBJECT: There is insufficient information in the neurosurgical literature regarding the long thoracic nerve (LTN). Many neurosurgical procedures necessitate a thorough understanding of this nerve's anatomy, for example, brachial plexus exploration/repair, passes for ventriculoperitoneal shunt placement, pleural placement of a ventriculopleural shunt, and scalenotomy. In the present study the authors seek to elucidate further the surgical anatomy of this structure. METHODS: Eighteen cadaveric sides were dissected of the LTN, anatomical relationships were observed, and measurements were obtained between it and surrounding osseous landmarks. The LTN had a mean length of 27 +/- 4.5 cm (mean +/- standard deviation) and a mean diameter of 3 +/- 2.5 mm. The distance from the angle of the mandible to the most proximal portion of the LTN was a mean of 6 +/- 1.1 cm. The distance from this proximal portion of the LTN to the carotid tubercle was a mean of 3.3 +/- 2 cm. The LTN was located a mean 2.8 cm posterior to the clavicle. In 61% of all sides the C-7 component of the LTN joined the C-5 and C-6 components of the LTN at the level of the second rib posterior to the axillary artery. In one right-sided specimen the C-5 component directly innervated the upper two digitations of the serratus anterior muscle rather than joining the C-6 and C-7 parts of this nerve. The LTN traveled posterior to the axillary vessels and trunks of the brachial plexus in all specimens. It lay between the middle and posterior scalene muscles in 56% of sides. In 11% of sides the C-5 and C-6 components of the LTN traveled through the middle scalene muscle and then combined with the C-7 contribution. In two sides, all contributions to the LTN were situated between the middle scalene muscle and brachial plexus and thus did not travel through any muscle. The C-7 contribution to the LTN was always located anterior to the middle scalene muscle. In all specimens the LTN was found within the axillary sheath superior to the clavicle. Distally, the LTN lay a mean of 15 +/- 3.4 cm lateral to the jugular notch and a mean of 22 +/- 4.2 cm lateral to the xiphoid process of the sternum. CONCLUSIONS: The neurosurgeon should have knowledge of the topography of the LTN. The results of the present study will allow the surgeon to better localize this structure superior and inferior to the clavicle and decrease morbidity following invasive procedures.