This doctoral thesis examines the value of information in settings, where two or more agents interact. In such situations, contrary to one-person decision problems, a more informative signal is not necessarily more valuable, and it may be profit-maximizing for an information seller to deliberately garble or damage his signal before selling it to another agent. More generally, the value of information depends on the precise contractual arrangement under which the information is to be transferred and used. I examine the following applications: (i) value of information in portfolio decision problems; and (ii) the transfer of information to a wealth-constrained investor. In a multiagent setting I examine (iii) the value of hared information services; and (iv) the value of information and flexibility for screening a heterogeneous consumer base.