Although the prevalence of HIV/AIDS and food insecurity was relatively high in Rufiji District in the mid-2000s, the extent to which the two problems were linked was empirically unknown. Therefore, a research for this thesis was done with the ultimate objective to determine the linkage between HIV/AIDS and food security at the household level. The main indicator of HIV/AIDS was a household having lost an adult member due to AIDS from January 2003 to December 2005: the main indicator of food security was dietary energy consumed (DEC) per adult equivalent per day. Data were collected among 225 households between November 2005 and October 2006 through Participatory Rural Appraisal (PRA), Household Income and Expenditure Survey (HIES) and structured interviews. Binary logistic regression was used for analysis and the dependent variable was food security in terms of food insecure (0) and food secure (I) based on kilocalories consumed per adult equivalent per day. The independent variables included having been affected by HIV/AIDS in terms of not affected (0) and affected (I). The results reveal that the odds for households affected by HIV/AIDS to be food secure were 0.705 times as high as the odds for households not affected by HIV/AIDS to be food secure. This means that households affected by HIV/AIDS were less likely to be food secure in comparison with those not affected by HIV/AIDS. The B statistic for having been affected by HIV/AIDS was negative (B = -0.350) meaning that being affected by HIV/AIDS had negative impact small (0.251) and not significant (p = 0.617) implying little impact of HIV/AIDS on food security. Based on these findings, it is concluded that although being affected by HIV/AIDS has negative impact on food security, it does not automatically make households food insecure, especially in a short run, and that some non-HIV/AIDS factors have bigger impact than that of HIV/AIDS on food security. On the basis of the on food security. However, the Wald statistic that shows the magnitude of impact was conclusion, it is recommended that efforts to improve food security among households affected by HIV/AIDS should consider both HIV/AIDS and non-HIV/AIDS factors.Rockefeller Foundation