For cultural, technical, and financial reasons, field sciences such as geology, ecology, and archaeology have lagged behind many of the laboratory sciences in making research data and samples available to the broader research community - but it is time for this to change, Marcia McNutt, Brian Nosek and colleagues emphasize in this Policy Forum. They note that data sharing is key for transparency and reproducibility. While they acknowledge that objects of study can often be fleeting, exceptionally rare, or forever changing in field sciences, there are numerous ways to better collect and share this data, they say. Some governments and journals have adopted a top-down approach, for example by requiring that data and samples be deposited upon publication. This may be a good first step, but they must also have the resources to verify the data or make it accessible, the authors say. They propose discipline-specific repositories that can provide more appropriate services. As well, machine-readable, quality-controlled, public metadata would better help researchers find, understand, and use the resources. Digitizing samples and other metadata will help scientists access the data remotely, no matter where they are located. "The days when scientists held on to samples and data hoping to squeeze out one more publication are ending," they say. They suggest that collaborative efforts and more data sharing will benefit everyone, highlighting, for example, the citation advantage associated with papers with open data, which suggests that stakeholders who promote transparency and reproducibility help themselves.