Scuba diving is becoming one of the favorite activities in marine leisure for human in recent years. However, increasing in diving frequency might result in substantial impacts on the marine environment, in addition, this might also influence on tourism quality (tourism perception). In this study, ecological and social carrying capacity approaches were performed to explore the tourism carrying capacity at two diving sites (Chaojing, CJ and Longdong, LD) in the north coast of Taiwan. Monthly number of divers were estimated from recorded number of divers in two days of weekends between July and September 2017 and May 2018. Number of colonies, coverage rate and damaged colonies of reefed corals were calculated by quadrat photos along four transects at two diving sites in September 2017 and May 2018. The perceived crowding of SCUBA divers were evaluated by a series of photographs (4 levels of diver numbers with 4 levels of proximity in 16 combinations) and questionnaires from June 2017 to May 2018. The results showed that the lowest number of dives was in May at both diving sites (4100 and 4900 dives for CJ and LD, respectively), while the highest number of dives was in August or September (5100 and 8100 dives for CJ and LD, respectively). For the same month, the number of dives in LD was higher than those in CJ. The number of colonies and coverage rate of corals were dominant by the encrusting corals, while damaged coral colonies were dominant by the branching corals for both diving sites. Coverage rate of corals was higher in May (36.8% and 28.0% for CJ and LD, respectively), while percent of damaged coral colonies was higher in September (6.4% and 17.1% for CJ and LD, respectively) for both diving sites. For the same month, coverage rate of corals in CJ was higher than that in LD, while percent of damaged coral colonies in LD was higher than that in CJ. The relationship between percent of damaged coral colonies and number of dives could be estimated for both diving sites. For CJ diving site, if the target for percent of damaged coral colonies was 10%, then the number of dives should be approximately 5000 dives. For LD diving site, if the target for percent of damaged coral colonies was 15%, then the number of dives should be approximately 7000 dives. A total of 257 respondents were analyzed in this study the respondent were dominant by males (>65%), diving here for more than 4 times (>68%), and diving with more than 5 divers (>49%). However, there was more first-time diving here for LD. The relationship between number of divers and proximity on perceived crowding was analyzed by multiple regression. The results showed that the number of divers was positive effect, while the proximity was negative effect on perceived crowding. Additionally, the number of divers was the most influential factor for divers’ perceived crowding for both diving sites. The divers would perceive crowding at a threshold of 5 divers for both diving sites. Based on the above results, the suitable dives number is 350 dives per day in weekends for CJ, and 500 dives for LD. If a target of 5 divers for each diving group is considered, thus there could be 70 groups and 20 groups per day for weekends and for weekdays, respectively in CJ, and 100 groups and 30 groups per day for weekends and for weekdays, respectively in LD. These results could be useful when considering recreational diving management measures in the north coast of Taiwan in future.