Innovations of the information age have made global-scale, high-resolution data more accessible to the scientific community than ever before. Such technological advances are timely in the face of rapid global change when research to enhance global stewardship and sustainability will increasingly rely on analyses of data from disparate sources. However, ecological datasets are typically characterized by sampling biases (e.g., nonrandomly missing data), hidden interactions, autocorrelations, and non-linearities among large numbers of possible explanatory variables. These traits are the norm rather than the exception for data describing natural systems, and combine to present formidable computational barriers to the application of standard analytical tools. Machine learning methods developed in the computer sciences offer an innovative solution to these computational barriers and are being applied with increasing frequency in ecological research. In contrast to traditional approaches that typically come with strong parametric assumptions, machine learning methods begin with empirical data and apply the axioms of probability to ‘let the data speak for themselves’. Machine learning algorithms accommodate many different types of data (categorical, continuous, etc.), sampling biases, and data incompleteness (e.g., nonrandomly missing data); they exhibit superior performance on high dimensional data (numerous potential explanatory variables) and employ methods to avoid over-fitting of the learning algorithms. Perhaps most attractively, machine learning algorithms achieve high-accuracy prediction by learning from non-linearities and complex interactions inherent in empirical data, and can often present results in easily interpretable outputs. The application of machine learning methods offers tremendous potential to revolutionize scientific investigation of patterns and processes in nature. However, many of these methods are unknown to the ecological community. Our main goal in this session is to bring together an international group of scientists investigating diverse biological problems through the innovative application of machine learning tools. Speakers in this session will increase awareness through a series of related case studies by showcasing the utility of machine learning algorithms to successfully address complex and often previously intractable biological questions. Examples include the application of machine learning algorithms to study fish dispersal ecology; assessing risk of invasive species; elucidating the complex ecology of migratory birds that often preclude standard ecological observation; and other intriguing examples of how machine learning methods can be applied to outstanding ecological questions.