Sampling decisions must emphasize not only data collection and analysis, but also data use in decisions made for protection and management of aquatic ecosystems. Although data gathering is often the main focus of an investigation, it only provides the opportunity for generating information. The quality of this information is dependent upon the method of data acquisition and analysis, and the effectiveness of the decision made depends on the entire process–not only sampling. The basic problem with analyzing aquatic ecosystems is their complexity which may not be adequately displayed if the sampling program is not carefully designed. Furthermore, even the most effectively designed program may not achieve the desired objectives if the sampling program design does not recognize the way the information will be used. This paper discusses the limitation of common experimental designs and sampling methods including the use of structural and functional measures, the sampling of natural and artificial substrates, and use of appropriate statistical tests. Certain sampling regimes, including sampling over a single annual cycle, may underestimate ecosystem variability. The use of artificial substrates for collections may be misleading if the behavior of the substrate over time is not understood. Nevertheless, artifical substrates may show greater replicability and reliability than collections from naturally heterogeneous substrates. Generators of information should understand the uses that will be made of the data and that the limitations of the data should be understood by those who must make decisions with it. A simple checklist is provided for use of investigators to ensure adequate preparation of hypotheses, selection of sampling methods, and use of statistical tests.