In this dissertation, I explore how economic agents conduct their investment decisions under uncertainty. Each of the three chapters empirically tests predictions from real-options frameworks of investment under uncertainty, shedding light on novel dimensions of agents' investment responses to uncertainty. In the first chapter, I study how the startup ecosystem responds to uncertainty. In the second chapter, I empirically measure the international transmission of uncertainty by examining US firms' investment responses to uncertainty induced by the 2016 Brexit Referendum. In the third chapter, I examine how uncertainty affects not only the level, but also the composition, of firms' capital stock using data on global shipping firms' investment and disinvestment decisions. Chapter 1 shows that economic uncertainty boosts dynamism among US startups. I introduce news- and survey-based measures of startup-relevant uncertainty and find that uncertainty is associated with net firm creation, and net job creation among young firms. I identify the channel by demonstrating, in a real-options framework, that venture capitalists (VCs) adjust their portfolios to take advantage of uncertainty. In contrast to mature firms delaying investment when facing uncertainty, VCs increase their investment spending during periods of heightened uncertainty, but do so by funding a large number of startups at low valuations. Critically, these dynamics play out solely at the earliest funding stages, implying greater experimentation by VCs. Buoyed by increased VC funding, startups accelerate their investment in technology and labor, producing more innovation and gaining greater traction. Looking at eventual outcomes, I provide evidence that startups receiving funding during high uncertainty periods are more likely to either fail or have exits with high multiples. My results point to uncertainty playing an important role in spurring "creative destruction" by stimulating risky startup activity in the economy. Chapter 2 (joint with Murillo Campello, Gustavo S. Cortes, and Fabricio D'Almeida) shows that the 2016 Brexit Referendum led American corporations to cut jobs and investment within US borders. Using establishment-level data, we document that these effects were modulated by the degree of reversibility of capital and labor. American job losses were particularly pronounced in industries with less skilled and more unionized workers. UK-exposed firms with less redeployable capital and high input-offshoring dependence cut investment the most. Data on the near-universe of US establishments also point to measurable, negative effects on establishment turnover (openings and closings). Our results demonstrate how foreign-born political uncertainty is transmitted across international borders, shaping domestic capital formation and labor allocation. Chapter 3 (joint with Murillo Campello and Hyunseob Kim) studies how economic uncertainty affects corporate asset composition and productivity using near-universe data on shipping firms' new orders, secondary-market transactions, and demolition of ships. Using a real-options framework, we show that shipping firms curtail both the acquisition and disposal of ships in response to heightened uncertainty. Critically, this mechanism operates primarily through cuts in new ship orders and demolition of older vessels --- decisions that are harder to reverse vis-a-vis deals in the used ship market. We use the escalation in Somali pirate attacks from 2009--2011 as a plausibly exogenous shock to uncertainty and find consistent results. The dynamics we identify are more pronounced when secondary ship markets are less liquid, as firms face stronger incentives to delay their decisions. Our results are novel in showing that uncertainty hampers "creative destruction" among mature firms in which these firms adopt technological innovation embodied in newer capital and dispose of old-vintage capital.