The purpose of the present study was to examine the relationship between exercise group characteristics and participation intention in elderly individuals via conjoint analysis. A total of 977 individuals (591 males and 386 females) aged ≥60 years were enrolled in the study. The included exercise groups comprised five factors: 1) instructor (professionals, university students, or volunteers); 2) place (1, 5, or 10 km); 3) expected effect (maintaining or increasing physical fitness, preventing dementia, or relieving stress); 4) characteristics (interaction, gaming, or self-pace); and 5) fee (free, 500 yen, or 1,000 yen). A total of 18 exercise groups were created by the orthogonal table. The choice-based conjoint comprised nine choice tasks. Among these, participants were required to choose between two exercise groups or no-choice option. The utility was higher for each factor as follows: 1) professional instructor, 2) closer place, 3) expected maintenance or increased physical fitness, 4) self-pace, and 5) free. The degree of each factor’s utility differed according to sex or exercise habits. Specifically, females had higher utility for university students than males, and those who had regular exercise habits had higher utility for professional instructors than those who did not. The results of the present study may help develop strategies to stimulate elder individuals in exercise-group participation.