The management of cotton cultivation and agriculture in general, depend on quality data enabling the retrieving of useful information as an aid in decision making related to management techniques and farm management . Part of this task depends intelligible data generated through the processes that make up this segment . Creating means for enabling the classification data is the starting point for making decisions regarding any corrections or adjustments in the mass data . The heterogeneity of data structures in farming systems create gaps in the sharing and integration of management systems. Generally, problem is the vertical integration of these solutions , currently the integration of databases in applications such as business intelligence , data warehousing and enterprise resource planning or ERP illustrate the problem . Although there are computational solutions guided consolidated the concepts above , there are still difficulties in terms of efficiency due to insignificant or incomplete data . The research proposes dimensions of data quality for qualifying masses of data with the specific farm . Despite the clarity of the need for information technology resources to improve management in agriculture in general there are relatively few initiatives developed or under development for this purpose . In general the requirements used in this process are generic and do not include the necessary particulars in agriculture . The aim of this research is to map the main requirements of data quality in agriculture and in particular in the production of cotton fiber in Brazil , providing subsidies to establish dimensions of data needed to measure data quality in this segment . The research focus is about the specification of dimensions for analysis of quality data specific to cotton production , the proposal is to establish qualitative parameters to determine actions enrichment data when necessary . The specification of criteria for qualification of dimensions used depends entirely on the context in which it is bounded on the research, metrics can take different amounts in different context . The absence of requirements for defining data structures can be seen in other dimensions like fill and duplication columns. There is no specific dimensions for measuring data quality in agriculture , dimensions may take different amounts depending on its context , the research presents a solution with the purpose of contributing to the qualification of masses of data and hence with possible enrichments . The contribution presented will create criteria to provide a methodology for qualification data given the particularities in this segment in Brazil , contributing complementary expert systems to generate information .