Legislation and company directives to end the use of the gestation stall for reasons involving sow welfare are in effect. There is conflicting evidence regarding the effects of sows maintained in stalls or when mixed into groups on reproductive performance and animal welfare. From the industry perspective, the housing of sows in stalls lowers the public’s perception of the welfare status of sows in the care of producers. From a producer perspective, mixing is problematic because it can cause pregnancy losses, reduced litter size, and loss of animals from fighting. This study was performed to evaluate the effects of day of mixing sows into groups during gestation on reproductive measures and measures of sow welfare. The treatments were applied in replicates during the summer months to obtain greater sensitivity to stress. The study was performed on a large 6000 sow commercial research farm in the Midwest. Sows of mixed parity (2 to 6) were weaned and checked for estrus. Upon detection of estrus, sows were assigned to treatment and bred by AI. Sows (n=1436) were assigned to housing treatment in: 1) stalls from weaning through gestation (Stall); 2) stalls from weaning until mixing at d 3-7 (D3 Mix); 3) stalls from weaning until mixing at d 13-17 (D14 Mix); and 4) stalls from weaning until mixing after d 35 (D35 Mix). All mixed sows were mixed into pens in a group of 58 sows. Each pen provided adequate floor space and each had one electronic sow feeding station. The Stall and the D35 Mix each served as controls to test for the effect of early mixing before pregnancy establishment at D3 Mix and later at time of implantation at the D 14 Mix. We recorded reproductive measures as pregnancy rate at day 30, farrowing rate, litter size, and for longevity as the proportion of sows that farrowed that remained in the herd to be bred within 10 days of weaning. Measures of well-being were also obtained and included early measures (Period 1) and later measures (Period 2). In Period 1, we recorded fighting events, lesions, and lameness, cortisol change, and body condition in the first 12 d after mixing or their movement into their permanent stall. In period 2, we recorded lesions, lameness and body condition from d 13 until farrowing. From a reproductive standpoint, conception rates were lower with the early D3 (87.1%) and D14 Mix (89.2%) treatments compared to the later D35 Mix (92.2%) and the Stall (96.2%) treatments. Farrowing rates remained lower in the D3 Mix (82.8%) compared to other treatments but the D14 Mix (87.8%) did not differ from the D35 Mix (90.5%) and the sows housed in Stall (92.8%). Litter size was not affected by any treatment (P > 0.10) and averaged 12.0 total born pigs. For sows that farrowed, there was an effect of treatment on the proportion of sows that were bred within 10 d of weaning. Fewer sows were bred (P < 0.05) for sows in the D3 and D14 Mix treatments compared to those in Stalls but the D35 Mix did not differ from either. The number of fights in the first 24 h after mixing was lower in the D14 Mix compared to the D3 and D35 Mix groups (P <0.0001) and did not occur in Stall. In period 1, cortisol was measured as the stress hormone, and cortisol increases were greatest (P< 0.05) in sows that were mixed compared to sows housed in the Stall. There were also (P < 0.05) effects of treatment, period and interactions for lameness, leg inflammation and lesions. In periods 1 and 2, mixing resulted in increased incidence of lameness, and increased lesion scores compared to those sows maintained in the Stall. Incidence of leg inflammation was not different in period 1 but was increased in period 2 for D3 Mix and Stall compared to other treatments (P < 0.05). A ranking from best to worst was performed using all measures for reproduction and well-being measures in each period. Final ranking order was the same for all measures and periods with the best ranking order: 1) Stall; 2) D35 Mix; 3) D14 Mix, and 4) D3 Mix. These results suggest that optimal reproduction and well-being can be achieved with use of stalls and that day of mixing can reduce all measures. Mixing in the first week results in reduced farrowing and well-being measures compared to mixing after the fifth week while mixing after the 2nd week shows intermediate effects. When mixing sows, short term responses for well-being and long term measures for reproduction and well-being must be considered to evaluate the effects of housing management. It would appear that strategic use of the use of the stall could be helpful for improving sow reproduction and welfare. Overall however, when examining the endpoint measures of reproduction and taking into consideration the classification of average well-being scores, our data suggests that in well-managed farms problems associated with day of mixing could have minimal impact.