Objective Early diagnosis is a key factor in the management of chronic obstructive pulmonary disease (COPD). Although mass screening is widely used, little is known about its accuracy and efficacy. This study investigated whether using spirometry during mass screening to detect COPD among community residents might be ineffective because of variability in the training and experience of examiners. Participants and Methods Both spirometry and a self-written questionnaire-based survey, including questions designed to detect respiratory symptoms, were conducted on community residents. Two separate studies were conducted on islanders living in similar environments. Study I was performed from 2004 to 2007 on Hachijyo Island residents, while study II, with a similar study design, was performed in 2003 on Inno Island residents. Results In study I, 3,592 subjects underwent examination over the 4-year study period; of these, 378 subjects underwent repeated examinations. Approximately 25% of the subjects had respiratory symptoms. Acceptable spirometry recordings were obtained for 62.0% (2004) to 84.1% (2006) of the subjects. In study II, 167 of the 254 subjects (65.7%) had respiratory symptoms. Acceptable assessment recordings were achieved in 254 subjects (95.5%). The suitability of the recordings was influenced by the extent/level of training of the examiners and the accompanying thoracic specialists. Conclusion We concluded that the effectiveness of health check-ups for COPD evaluation using spirometry was greatly influenced by the quality of the examiners, even when the subjects had respiratory symptoms. Thus, we recommend caution when screening for early signs of COPD during health check-ups.