Data sharing is the practice of making data available for use by others. Ecologists are increasingly generating and sharing an immense volume of data. Such data may serve to augment existing data collections and can be used for synthesis efforts such as meta-analysis, for parameterizing models, and for verifying research results (i.e., study reproducibility). Large volumes of ecological data may be readily available through institutions or data repositories that are the most comprehensive available and can serve as the core of ecological analysis. Ecological data are also employed outside the research context and are used for decision-making, natural resource management, education, and other purposes. Data sharing has a long history in many domains such as oceanography and the biodiversity sciences (e.g., taxonomic data and museum specimens), but has emerged relatively recently in the ecological sciences. A review of several of the large international and national ecological research programs that have emerged since the mid-1900s highlights the initial failures and more recent successes as well as the underlying causes-from a near absence of effective policies to the emergence of community and data sharing policies coupled with the development and adoption of data and metadata standards and enabling tools. Sociocultural change and the move towards more open science have evolved more rapidly over the past two decades in response to new requirements set forth by governmental organizations, publishers and professional societies. As the scientific culture has changed so has the cyberinfrastructure landscape. The introduction of community-based data repositories, data and metadata standards, software tools, persistent identifiers, and federated search and discovery have all helped promulgate data sharing. Nevertheless, there are many challenges and opportunities especially as we move towards more open science. Cyberinfrastructure challenges include a paucity of easy-to-use metadata management systems, significant difficulties in assessing data quality and provenance, and an absence of analytical and visualization approaches that facilitate data integration and harmonization. Challenges and opportunities abound in the sociocultural arena where funders, researchers, and publishers all have a stake in clarifying policies, roles and responsibilities, as well as in incentivizing data sharing. A set of best practices and examples of software tools are presented that can enable research transparency, reproducibility and new knowledge by facilitating idea generation, research planning, data management and the dissemination of data and results.