Knowledge in society is often dispersed, with different individuals holding different pieces of information. If aggregated correctly, this information can beat that of any single individual, an idea captured by the expression “wisdom of the crowd”. There is considerable interest in harnessing this wisdom for forecasting and decision-making. Prediction markets are popular tools to “crowd-source” distributed information for forecasts. Such forecasts can be very valuable for decision makers. Commercial companies, for instance, can benefit tremendously from accurate forecasts regarding the future demand for their products. Many decision-making problems, however, require more than just a peek into the future - they require conditional forecasts. To decide, for instance, between alternative marketing campaigns, a company needs to understand how each of the alternatives will affect sales. Finding mechanisms that properly incentivise participants to provide their information for such conditional forecasts is non-trivial, but can be done through so-called decision markets. Decision markets have only recently been described in theory. Aim of this project is to investigate fundamental theoretical properties of decision markets and their relation with alternative decision-making mechanisms such as voting; to provide a proof-of-concept regarding their functioning in human-subject experiments; and to test them in practical applications.