Breast feeding improves the health of mothers and infants; the UK has low rates, with marked socioeconomic inequalities. While trials of peer support services have been effective in some settings, UK trials have not improved breast feeding rates. Qualitative research suggests that many women are alienated by the focus on breast feeding. We propose a change from breast feeding-focused interactions to respecting a womans feeding choices, inclusion of behaviour change theory and an increased intensity of contacts in the 2 weeks after birth when many women cease to breast feed. This will take place alongside an assets-based approach that focuses on the positive capability of individuals, their social networks and communities.We propose a feasibility study for a multicentre randomised controlled trial of the Assets feeding help Before and After birth (ABA) infant feeding service versus usual care. A two-arm, non-blinded randomised feasibility study will be conducted in two UK localities. Women expecting their first baby will be eligible, regardless of feeding intention. The ABA infant feeding intervention will apply a proactive, assets-based, woman-centred, non-judgemental approach, delivered antenatally and postnatally tailored through face-to-face contacts, telephone and SMS texts. Outcomes will test the feasibility of delivering the intervention with recommended intensity and duration to disadvantaged women; acceptability to women, feeding helpers and professionals; and feasibility of a future randomised controlled trial (RCT), detailing recruitment rates, willingness to be randomised, follow-up rates at 3 days, 8 weeks and 6 months, and level of outcome completion. Outcomes of the proposed full trial will also be collected. Mixed methods will include qualitative interviews with women/partners, feeding helpers and health service staff; feeding helper logs; and review of audio-recorded helper-women interactions to assess intervention fidelity. Study results will inform the design of a larger multicentre RCT. The National Research Ethics Service Committee approved the study protocol. ISRCTN14760978; Pre-results.