It is important to better understand why in Mozambique, a land abundant country by Sub-Saharan Africa countries standards, there are seemingly many land-poor households. In Mozambique, recent studies show that land plays a key role in generating income for rural households, and those with less land tend to be poor. Thus, the present study was intended to further confirm the existence of land-and income-poor households, to understand the nature and evolution of this group and to clarify the factors that are leading to its creation. Data from 521 rural households across the cotton belt in Northern Mozambique and from the Ministry of Agriculture 1996 National Smallholder Survey were used. Additional data collection consisted of in-depth case studies in selected villages of Nampula and Cabo Delgado Provinces, and field area measurements in these locations. This study has confirmed the existence of a significant group of land- and income-poor households in Northern Mozambique. Results also showed that land inequality throughout the country is similar to that found in Nampula and Cabo Delgado, and this inequality has changed little three years after the ending of the war. Analysis of the relationship between land holdings and household income showed that the observed inequality in land distribution is a problem, in that land holdings are a major determinant of household income. Thus, land poor households face both a serious income shortage and, in all likelihood, a critical food security problem. Results show that understanding the way the customary tenure system and related traditional decision making structures operate within the society is important to better understand why some smallholders are land-poor. Findings indicate that there is an effect on land access of local social hierarchy. In the Macua society specifically, findings suggest that there is an effect of local social hierarchy on land access. The role of the principal Atata was shown to be especially important in the Nampula villages. It was shown that the size of Matala under the control each of principal Atata varies greatly in these villages; that per capita land availability within these Matala also varies greatly; and that households in small Matala are only slightly more likely to possess fields outside their Matala than households in large Matalas. We, therefore, hypothesize that the size of Matala is a key determinant of total land availability at the household level, and suggest focusing further research efforts on this issue.