The objective of this Thesis is to explore several related questions with regards to criteria for viability in scalar field theories. Roughly the first half is devoted to theoretical criteria, while the second half focuses on phenomenological ones. We begin with an overview of theories that violate the null energy condition, highlighting the pathologies that inevitably appear. We then present a theory that violates the null energy condition while remaining free of the problems that plagued previous attempts. Next we explore a global condition for classical stability in scalar field theories, namely, the requirement that the total energy of the space-time be positive. This property is guaranteed if the theory admits a positive energy theorem. After reviewing existing proofs of positive energy for canonical scalar fields, we then extend those proofs to theories with derivative interactions, proving a positive energy theorem for a wide class of P(X) theories. The second half of this Thesis considers experimental constraints on scalar field theories. We focus on what may be learned from atom interferometry experiments, which have been a powerful probe of fundamental physics for over two decades but only recently gained the ability to constrain screened scalar field theories. We present a general analytic and numerical framework for precise predictions of scalar field theories in atom interferometry experiments, and use those techniques to derive new limits on chameleon and symmetron field theories.