BACKGROUND: Thumb-base osteoarthritis is a common degenerative condition that produces symptoms including pain and reduced hand function. Trapeziectomy is the mainstay surgical intervention following failure of conservative management and steroid injection, with carpometacarpal (CMC) joint replacement anticipating faster recovery time and return to work. However, there is limited evidence on recovery following both surgeries, which affects patient counselling and future research design. METHODS: A preregistered (CRD42023450865) systematic review of the literature was conducted according to the PRISMA guidelines. Disability of the Arm, Shoulder and Hand (DASH) questionnaire scores, grip strength, and key pinch strength scores were obtained at different timepoints during follow-up. Standardized mean changes (SMC) were calculated and pooled through random-effects meta-analysis before being plotted at monthly timepoints. RESULTS: Of 63 articles eligible for inclusion, 25 studies were included in the meta-analysis. Subgroup analyses were conducted comparing types of trapeziectomy, joint replacement implants, and immobilization protocols. At the 1-month follow-up, joint replacement was associated with a trend of improvement in the DASH score, whereas trapeziectomy was associated with a deterioration. At 12 months, joint replacement showed a trend of greater improvement in hand function than trapeziectomy. Key pinch strength returned to baseline function at 1.5 months after joint replacement, compared with 6 months after trapeziectomy. Newer dual-mobility implants were associated with a quicker trend of recovery in DASH and key pinch strength than single-mobility implants. However, confidence intervals overlapped considerably, and most study arms had a high risk of bias, so these findings should not be considered statistically significant. CONCLUSION: This study shows a trend towards faster early recovery after CMC joint replacement compared with trapeziectomy, although the evidence remains limited. Randomized clinical trials with systems to frequently capture patient-reported outcomes data are required to study the temporal course of recovery for both surgeries.