This study sought to identify factors associated with objectively assessed exercise behavior in Chinese patients with early-stage cancer. Three hundred and fifty one cancer patients were recruited from the Affiliated Jiangyin Hospital of Southeast University Medical College and the Nantong Tumor Hospital. One-way ANOVA, Pearson Chi-square tests and regression analysis were employed to identify the correlations between physical exercise and the measured factors. The results showed that occupation type (χ2 = 14.065; p = 0.029), monthly individual monthly income level (χ2 = 24.795; p = 0.003), BMI (χ2 = 15.709; p = 0.015) and diagnosis (χ2 = 42.442; p < 0.000) were significantly correlated with the subjects self-reported exercise with different frequency per week. Differences in the frequency of exercise were associated with different degrees of reported Benefit Finding (BF) (F = 24.651; p < 0.000), communication with doctors (F = 15.285; p < 0.000), medical coping modes (F = 45.912; p < 0.000), social support (F = 2.938; p = 0.030), depression (F = 6.017; p < 0.000), and quality of life (F = 12.288; p < 0.000). Multiple regression analysis showed that 1.6%-6.4% of the variance in five variables, excluding social support and optimism could be explained by exercise. Our results indicated that benefit finding, medical coping modes, communication with doctors, social support, depression and quality of life were significantly correlated with exercise. The variance in several psychosocial factors (benefit finding, medical coping modes, the communication with doctors, depression and quality of life) could be explained by exercise. Psychosocial factors should be addressed and examined over time when evaluating the effect of physical exercise that is prescribed as a clinically relevant treatment.