Musculoskeletal disorders are common in subjects with type 2 diabetes mellitus (T2DM) and may represent a barrier to physical activity (PA). In the Italian Diabetes and Exercise Study, we assessed whether a strategy aimed at minimizing the risk of injury and worsening of musculoskeletal symptoms favors participation in exercise training and results in improvements in joint-specific functional scores, along with amelioration of metabolic profile. Sedentary patients with T2DM (n = 606) from 22 outpatient diabetes clinics were randomized to twice-a-week supervised aerobic and resistance training plus exercise counseling (EXE) versus counseling alone (CON) for 12 months. Musculoskeletal symptoms were evaluated at baseline in the entire cohort by the use of a self-reported questionnaire in order to design tailored exercise programs with exclusion of painful segments, and validated joint-specific scores were calculated at the end of study in a subgroup of CON (n = 65) and EXE subjects (n = 62). Musculoskeletal symptoms were present in more than ¾ of EXE and CON subjects. Yet, volumes of non-supervised PA were high, whereas dropout rates and adverse events were low in both groups. Moreover, in the EXE group, no difference was observed between patients with and without symptoms in session attendance and PA/exercise volume. Scores for symptoms and functional status of limbs and spine were significantly better in EXE than in CON subjects and correlated with PA/exercise volume and improvements in fitness parameters. Preliminary evaluation of musculoskeletal symptoms is useful in favoring compliance with supervised training programs and obtaining significant benefits to the functional status of the involved joints.