Physical activity (PA) is essential for physical, mental, and emotional health; however, few adults engage in enough PA. Group exercise environments such as CrossFit can promote sustained exercise habits through social influence, support, and norms. This cross-sectional study aims to provide evidence for PA social influence at CrossFit. CrossFit members (n = 62) reported PA, workout logging frequency, and members at their gym they: (1) work out with and (2) go to with personal matters. Separate linear network autocorrelation models (LNAMs) determined if individuals reported similar PA scores as those of their social ties at CrossFit that they work out with and/or those they go to for personal matters. Participants reported a mean of 2740.55 MET minutes/week (SD = 1847.08), working out with a mean of 9.89 members (SD = 6.26), and speaking to a mean of 2.66 members about personal matters (SD = 3.68). A person’s PA was significantly associated with that of their ties they go to with personal matters (PEp = 0.08, SEp = 0.02), but was not associated with the PA of their ties they work out with (PEw = 0.02, SEw = 0.01). Social influence on PA levels was present when a deeper connection is made between members. Fostering and promoting deeper connections between members may help promote PA and continued exercise habits.