The greatest challenge to the welfare of dairy cows occurs in the peripartum period. Given the perception that cow welfare is better in more natural environments, it was hypothesised that cows in a PASTURE-based production system (cubicle housing with grass silage pre-partum and rotational grazing with concentrate supplementation post-partum) would have improved peripartal welfare compared to cows in a HOUSED production system (cubicle housing with a total mixed ration [TMR], pre-partum and post-partum). Blood samples were analysed for acute phase proteins (APP), cortisol, white blood cell (WBC) differential and counts and other biochemical metabolites as non-specific indicators of sub-clinical ill-health and nutritional stress. Daily monitoring of rectal temperature (RT) and rumen fill (RF) scores were used to monitor ill-health and nutritional status. Reproductive health and welfare (calving difficulty, retained placenta, puerperal metritis, endometritis and oestrous cyclicity) was also recorded. No differences were found between treatments for APP, cortisol or WBC. Blood metabolite differences indicated that PASTURE cows were under greater nutritional stress than HOUSED cows. HOUSED cows showed an increase in RF score from day 0 to 10 post-partum and had a higher RF score than PASTURE cows. PASTURE cows had an overall lower RT and lower incidence of reproductive disorders. Results primarily reflect nutritional differences between treatments with PASTURE cows showing greater potential nutritional/metabolic stress in early lactation which has attendant implications for welfare. Nevertheless, this did not result in inferior health and, in accordance with our hypothesis, PASTURE cows’ reproductive health and welfare tended to be better than that of HOUSED cows.