Psychological characteristics, such as depression, anxiety or negative illness perception are highly prevalent in patients with several types of OA. It is unclear whether there are differences in the clinical and psychological characteristics of patients with thumb carpometacarpal (CMC-1) osteoarthritis (OA) scheduled for nonsurgical treatment and those with surgical treatment. (1) What are the differences in baseline sociodemographic characteristics and clinical characteristics (including pain, hand function, and health-related quality of life) between patients with thumb CMC-1 OA scheduled for surgery and those treated nonoperatively? (2) What are the differences in psychological characteristics between patients scheduled for surgery and those treated nonsurgically, for treatment credibility, expectations, illness perception, pain catastrophizing, and anxiety and depression? (3) What is the relative contribution of baseline sociodemographic, clinical, and psychological characteristics to the probability of being scheduled for surgery? This was a cross-sectional study using observational data. Patients with CMC-1 OA completed outcome measures before undergoing either nonsurgical or surgical treatment. Between September 2017 and June 2018, 1273 patients were screened for eligibility. In total, 584 participants were included: 208 in the surgery group and 376 in the nonsurgery group. Baseline sociodemographic, clinical, and psychological characteristics were compared between groups, and a hierarchical logistic regression analysis was used to investigate the relative contribution of psychological characteristics to being scheduled for surgery, over and above clinical and sociodemographic variables. Baseline measures included pain, hand function, satisfaction with the patient's hand, health-related quality of life, treatment credibility and expectations, illness perception, pain catastrophizing, and anxiety and depression. Patients in the surgery group had longer symptom duration, more often a second opinion, higher pain, treatment credibility and expectations and worse hand function, satisfaction, HRQoL, illness perception and pain catastrophizing compared with the non-surgery group (effect sizes ranged from 0.20 to 1.20; p values ranged from < 0.001 to 0.044). After adjusting for sociodemographic, clinical, and psychological factors, we found that the following increased the probability of being scheduled for surgery: longer symptom duration (standardized odds ratio [SOR], 1.86; p = 0.004), second-opinion visit (SOR, 3.81; p = 0.027), lower satisfaction with the hand (SOR, 0.65; p = 0.004), higher treatment expectations (SOR, 5.04; p < 0.001), shorter perceived timeline (SOR, 0.70; p = 0.011), worse personal control (SOR, 0.57; p < 0.001) and emotional response (SOR, 1.40; p = 0.040). The hierarchical logistic regression analysis including sociodemographic, clinical, and psychological factors provided the highest area under the curve (sociodemographics alone: 0.663 [95% confidence interval 0.618 to 0.709]; sociodemographics and clinical: 0.750 [95% CI 0.708 to 0.791]; sociodemographics, clinical and psychological: 0.900 [95% CI 0.875 to 0.925]). Patients scheduled to undergo surgery for CMC-1 OA have a worse psychological profile than those scheduled for nonsurgical treatment. Our findings suggest that psychological characteristics should be considered during shared decision-making, and they might indicate if psychological interventions, training in coping strategies, and patient education are needed. Future studies should prospectively investigate the influence of psychological characteristics on the outcomes of patients with CMC-1 OA. Level III, therapeutic study.