The newsvendor problem aims to optimally choose a level of order quantity to respond to a known demand distribution with the objective of maximizing expected return. In practice, the decision maker is often challenged with more complex settings involving multiple decisions and uncertainties. For instance, firms may benefit from choosing the set of customer orders to satisfy. It may also be worthwhile for many firms to select a supply portfolio instead of relying on a single procurement mode. This paper provides novel optimization models and solution techniques that can help businesses to achieve the maximum performance from a given production system by optimally selecting customer demands, procurement quantity, spot market purchase and option contract usage. We specifically focus on the special case of normally distributed random variables, and provide an exact solution method. When the primary procurement quantity is not a decision variable, the problem becomes a version of a Stochastic Knapsack Problem. For this case, we present an efficient heuristic solution algorithm based on properties of an optimal solution and empirically show that it provides high-quality solutions. We also provide a broad numerical study to examine the sensitivity of integrated procurement and demand selection strategies to key problem parameters. (C) 2016 Elsevier B.V. All rights reserved.