Abstract Background and aim: Protection, promotion and support of infant and young child feeding contribute to the wellbeing of mother and child, social equality, socioeconomic development, food security and poverty reduction. The aim of this study is to identify different socio-demographic factors affecting breast feeding behaviour of mothers. This includes how long they breast fed their last born child, if they gave predominant feeding or breast milk few days after birth, when they gave complementary food and what kind of complementary food they gave. We also wanted to study how HIV affected the mothers. Material and method: This project was first part of the Tanzanian-Norwegian AIDS Project (MUTAN; 1990-1995) and later part of the NUFU (Norwegian Universities’ Committee for Development Research and Education) funded project, “Health Systems Research and Health Promotion in Relation to reproductive Health in Tanzania”. The cross sectional survey included a total of 1,528 subjects between 15-45 years. This present study included 666 women between 15-45 years. The women were mothers and had their last delivery between 1995 and 2005. Data was collected by questionnaires. Association between breast feeding duration and demographic factors was studied, using Cox survival analysis. Practises of predominant and complementary food’s relation to socio-demographic factors were studied using multiple logistic regressions. The main variables were; breast feeding duration (months), food given few days after birth (predominant food), timely introduction of complementary foods and type of complementary foods (fluid/solid). Results: A majority initiated breast feeding (98, 7%). Median breast feeding duration was 24 months. Most of the mothers who had terminated breast feeding applied that the child was old enough. The oldest women breast fed significantly longer than the youngest age group. Women who had their last delivery after 2002 breast fed significantly shorter than women who had their last delivery before 2003. Farmers, married/cohabiting women and women who had their last delivery between 2003 and 2005 were more likely to give breast milk few days after birth. HIV positive women and women who had their last delivery between 2003 and 2005 introduced complementary food before the child was 4 months old. Almost 85% of the mothers give solid food as complementary food. Children who were born between 2003 and 2005 and children who got predominant food, few days after birth had higher risk of getting fluids as complementary food. Conclusion: The study indicates that breast feeding prevalence in Oria is high. The practice of predominant feeding and the onset of complementary food is however not optimal. Young women, single mothers, employed mothers, HIV positive women, and women who gave birth after 2002 are groups that need extra support and protection in breast feeding interventions. Education is important to promote safe breast feeding and sexual behaviour. Paternal support is also essential to secure optimal breast feeding patterns.