Objectives: To investigate the efficacy of communication skills training (CST) on improving communication behaviors of employees, we conducted a randomized controlled trial. Methods: We randomly divided 128 white-collar workers to either a CST intervention group (n=64) or a nonintervention control group (n=64). The three-hour CST session was conducted by an occupational physician. A Likert scale was used to assess four aspects of communication behavior: conversing so as not to put pressure on the subject, displaying empathy and support, helping the subject sort out problems, and thinking together to solve problems. Communication behavior scores were compared between the intervention and control groups three months after training. Results: Intention-to-treat analyses using mixed effect models showed a significant interaction between group and time (p=0.045) for all participants for "thinking together to solve problems," with an effect size (Cohen's d) of 0.37. For the other three items, although an improving trend was noted in the intervention group, no statistically significant interactions between group and time were observed. The sub-group analysis results for the four items' lower half baseline value for the items "helping the subject sort out problems" and "thinking together to solve problems" showed a significant interaction between group and time (p < 0.01) with effect sizes of 0.67 and 1.04, respectively. Conclusion: Brief CST for workers may be useful in improving communication behavior between individual employees. Given that CST improves and contributes to developing positive mental health, we hope that further intervention studies will be conducted in the future.