Abstract: Objective To explore the trajectory and influencing factors of changes in physical activity (PA) in patients with non-small cell lung cancer (NSCLC) during chemotherapy. Methods A random sampling method was used to select NSCLC patients receiving chemotherapy in Suzhou Ninth People′s Hospital as the research objects. Patients′materials were collected using a general information questionnaire, the M. D. Anderson Symptom Inventory (MDASI), the Family Adaptation, Partnership, Growth, Affection, and Resolve index (APGAR), and the Self-Rated Approaches for Health Promotion Scale (SRAHP). The International Physical Activity Questionnaire-long-form (IPAQ-L) was used to assess patients′ PA levels before the first chemotherapy course (T0), before the second course (T1), before the fourth course (T2), and before the sixth course (T3). The latent growth mixture model (LGMM) was used to identify the trajectory types of patients′physical activity metabolic equivalents (MET). Unordered multinomial Logistic regression analysis was used to explore the influencing factors of PA change trajectories in NSCLC patients during chemotherapy. Results A total of 213 NSCLC patients completed this study. The PA values of NSCLC patients at T0, T1, T2 and T3 were (1 091.29±285.76), (456.43±92.78), (467.61±94.72), and (934.35±199.18) MET -min/week, respectively, with significant between-group differences (F=645.601, P < 0.001). After fitting with the LGMM, three latent categories were selected: the low PA maintenance group (accounting for 28.64%), the medium PA decline followed by late-stage recovery group (accounting for 51.64%), and the medium PA decline followed by mid-stage recovery group (accounting for 19.72%). Analysis showed that age, tumor stage, health behavior capacity, symptom severity and symptom distress, and family care were influencing factors for the PA trajectorycategories of NSCLC patients during chemotherapy (all P < 0.05). Analysis of variable interaction effects found that health behavior capacity (OR=0.376, 95% CI, 0.192 to 0.853) had a strong protective effect on the medium PA decline followed by mid-stage recovery group (there was an interaction effect, P < 0.05). Conclusion There is population heterogeneity in PA among NSCLC patients during chemotherapy. Medical staffs should develop targeted measures based on the population characteristics and influencing factors of patients′ PA trajectories to improve their PA levels.