Many of the countries in the Southern Africa region have very sketchy trade data, due to disruptions in data collection (caused by war in the case of Mozambique and sanctions-induced secrecy in South Africa, for instance) or weak statistical institutions. These trade data are also often in conflict with one another, occasionally even by several orders of magnitude. While much anecdotal information on trade flows exists (especially on newly created trade), it is only now that basic and systematic research on formal trade flows in Southern African is being conducted (e.g. USAID-funded EAGER project and Will Masters at Purdue University, David Evans at University of Sussex, International Food Policy Research Institute's Macroeconomic Reforms and Regional Integration in Southern Africa — MERRISA project). The sparse research agenda on trade issues is a serious problem, in light of Southern Africa's desire for deepened regional integration, improved prospects for economic growth, and equitable development. This paper will present trade (exports and imports, at a 1-digit SITC level, but with maize listed as a separate trade item) data from approximately 20 different sources, both from individual countries and international organizations, for the period 1980-present. The paper will show the major inconsistencies in the trade data between the different data sources. The main contribution of this paper is the reconciliation of the trade data, using a 'cross-entropy' estimation method, to make efficient use of prior information about the structure of trade in the Southern Africa region. The benefit of this approach is that it is very flexible and undemanding in its statistical assumptions, and it allows the use of the trade information in a variety of forms, such as inequality constraints, errors in measurement, and extraneous estimates of various parameters. To date, the cross-entropy method has been applied to the Social Accounting Matrices (SAMs) for the MERRISA countries, including the SAM for Zimbabwe.