We study sequential search without priors. Our interest lies in decision rules that are close to being optimal under any prior, at each point in time, and after each history. We call these rules dynamically robust. The search literature employs optimal rules based on cutoff strategies that are not dynamically robust. We uncover general principles of dynamically robust decision making, such as the history independence and the necessity to randomize. We then derive dynamically robust rules and show that their performance exceeds 1/4 of the optimum against all environments and 1/2 of the optimum against binary environments.