Bubbles are recurrent events, which contribute to both macroeconomic and employment volatility. We introduce stochastic bubbles in the standard search-and matching model of the labor market. The economy alternates between latent and bubbly states, each being associated with a distinct solution for the market value of firms (respectively, stable or explosive). Bubbles in firm value induce distortions in hiring decisions and wages, which we explicitly characterize. Faced with bubbles, the social planner optimally deviates from the standard Hosios efficiency condition. The optimal share of workers in total surplus must be above the elasticity of hiring rates, by a small but increasing amount as the bubble expands. Finally, our specification for bubbles significantly improves the quantitative ability of the model to match U.S. data, along both real and financial dimensions.