Background: Observational data demonstrate an inverse relationship between PA and Met to disease outcomes in CRC & BC pts. A mechanism that these could impact cancer recurrence and mortality is hypothesized to involve insulin and related growth factors. We investigated the effects of PA, Met or the combination on metabolic biomarkers in CRC & BC pts. Methods: In a phase 2 RCT, stage I-III CRC & BC survivors at least 2 months from completing standard therapy (excluding hormone rx or trastuzumab) were randomized to PA, Met, PA + Met or control. Major eligibility included absence of recurrence or diabetes (glucose < 160 (random) or < 126 (fasting) mg/dl) and exercising < 120 min/wk. The PA intervention consisted of supervised aerobic training at least 2 x /wk. Metformin dosing was 850 mg 1x/day titrated to 850 mg 2x/day after 2 wks if tolerated. Interventions were 12 weeks in duration. Fasting bloods at baseline & 12 wks were analyzed for insulin (1^oendpoint), leptin, IGF1, IGFBP1 & IGFBP3. Results: 139 pts were enrolled: 62% BC/38% CRC; 83% female; median BMI 28.3; median 2 yrs from dx; median age 56 (range 34-79). 107 pts completed assigned therapy. Pts in PA and PA + Met arms increased PA by 166 and 140 min/wk vs 30 min/wk in controls (both P<0.0001). Pts in the Met and PA + Met arms lost weight vs controls (-1.41 and -0.91 kg vs + 1.97 kg, both P <0.0001). Both interventions had impact on metabolic biomarkers (Table). Conclusions: PA and Met both led to significant changes in insulin and other biomarkers in CRC & BC survivors with potential synergistic effect on leptin with dual intervention. Clinical trial information: NCT01340300