Background <p>Lung cancer patients’ recovery is hindered by persistent symptoms, but the relationship between symptoms and physical activity is not well understood.</p> Objective <p>The objective of this study is to examine changes in symptoms and physical activity among lung cancer patients and their short-term predictive relationship.</p> Methods <p>Ninety-six lung cancer patients (average age 59.70) were selected. Symptoms (pain, fatigue, distress, cough, dyspnea, and sleepiness) were monitored three times daily from preoperative to postoperative day 6 using the Ecological Momentary Assessment. Step count was continuously monitored via a smartwatch. Short-term trends in symptoms and step count were analyzed. A hierarchical linear model was used to examine the immediate and lagged correlations between symptoms and step count, with confounders controlled.</p> Results <p>Symptoms and step counts showed fluctuation with a jagged changing trajectory. Step count decreased significantly when symptoms were worse than the individual average (<i>P</i> &lt; 0.001), except for cough. Symptoms were significantly reduced when step count significantly increased (<i>P</i> &lt; 0.001). Pain (<i>β</i> =  − 73.384, <i>P</i> &lt; 0.001) had a negative predictive effect on step count at the next moment, after controlling for step count. Step count at the same time had a significant effect on alleviating pain (<i>β</i> =  − 0.0002, <i>P</i> &lt; 0.05), fatigue (<i>β</i> =  − 0.0003, <i>P</i> &lt; 0.05), and sleepiness (<i>β</i> =  − 0.0003, <i>P</i> &lt; 0.001), next moment, after controlling for symptoms. There was a significant intrinsic predictive effect between symptoms and step count (<i>P</i> &lt; 0.001).</p> Conclusion <p>Lung cancer patients experience fluctuations in symptoms and step count during the perioperative period. There is a significant relationship between symptoms and step count on the same day. Future studies could explore interventions to improve symptoms and mobility.</p>