PURPOSE: Constructing a lateral key pinch (KP) is a universal aim of any functional upper limb surgery program for tetraplegia. Three stages are required: (1) activating the pinch mechanism by flexor pollicis longus tenodesis to the radius or by tendon transfer to the flexor pollicis longus, (2) simplifying the polyarticular chain, and (3) positioning the thumb column. We compared 2 techniques for accomplishing the latter stage, 1 utilizing arthrodesis of the carpometacarpal joint (CMC) and 1 that did not require arthrodesis of the CMC. MATERIALS AND METHODS: We reviewed 40 cases of KP reconstruction at a mean follow-up of 7.4 years: 17 who had undergone CMC arthrodesis and 23 without CMC arthrodesis. In this group, an abductor pollicis longus tenodesis was necessary to properly position the thumb column in 17 patients. RESULTS: Active KP cases with CMC arthrodesis were significantly stronger than those without an arthrodesis. For passive KP cases, the difference between those cases with CMC arthrodesis and those without was not significant. Regarding opening, for active KP cases with CMC preservation alone, the mean distance between the thumb pulp and the index finger was 4.0 cm at rest and 5.8 cm when passively grasping large objects; for active KP cases without arthrodesis, these values were 3.4 and 6.8 cm, respectively, with the wrist in flexion. For passive KP cases, these values were 2.2 and 3.5 cm with CMC arthrodesis compared with 2.4 and 6.9 cm without arthrodesis. Overall, 23.5% of patients with CMC arthrodesis could not maintain contact between the thumb and the index finger compared with 30.4% without arthrodesis. CONCLUSIONS: Active KP is stronger with than without CMC arthrodesis; however, the KP reconstruction does not open as far when grasping large objects. For passive KP, CMC arthrodesis significantly limits passive opening, with no gain in strength. Neither technique is superior in terms of KP stability. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic IV.