This study aimed to investigate the relationships among serious leisure, recreation specialization and leisure benefits. The study used a questionnaire-based survey by judgment sampling recreational scuba divers. A total of at 413 valid questionnaires were collected with valid rate of 97.9%. Through descriptive statistics, Pearson product moment correlation analysis and structural equation modeling analysis (SEM) to verify our model. The results showed a good fit to overall model, between the various dimensions of the impact of direct and indirect relationships and the dimensions have a significant correlation: (1) the serious leisure significantly directly positive influence the recreation specialization , flow experience and leisure benefits. (2) the recreation specialization significantly directly positive influence flow experience. (3) the flow experience significantly directly positive influence leisure benefits. (4) the recreation specialization significantly directly negative influence leisure benefits.