This paper discusses main problems associated with evaluation of performance and impact of long-term environmental programs. Lack of data standards, incompleteness of archived datasets and insufficient statistical power were identified as main limits in functionality of monitoring networks. To avoid these failures, environmental programs should be designed with inception to incorporate data management as their integral part. Especially in global programs, local and regional data managers should invest significant proportion of their effort to handle documentation in terms of standardized coding, data formats, metadata coding and consistency of records over time. Up-to-date trends in building knowledge-based infrastructures are illustrated using example of monitoring of atmospheric pollution by persistent organic pollutants (POPs). Conceptual model usable to facilitate the integration and analysis of data on POPs concentrations is introduced with its multilayer hierarchy of entities (POPs as nomenclature classes, couples “observation – measurement” as content classes). Robust set of statistical methods for processing of time series of concentration data is discussed from the viewpoint of practical implementation within running monitoring programs. It consists of the following components: baseline pollution estimates, uncertainty analyses, spatial extrapolations, effect size estimates, time trend identification and quantification. Development of tools supporting standardized environmental data management is rapidly expanding field of science which results in the following challenges for applied informatics and statistics: log-term sustainability of information systems, data-related metadata coding and archiving, tools for automated integration and reporting of data.