In this paper, we propose a novel method for robust bifurcation analysis of systems with dynamic uncertainties. First, we formulate a robust bifurcation analysis problem for parameter-dependent systems with norm-bounded uncertainties. Next, to solve this problem, we define a new concept of robust hyperbolicity of an equilibrium that for any uncertainty an uncertain linear system has no neutral pole and the number of unstable poles is constant. A necessary and sufficient condition for the robust hyperbolicity is derived from the Nyquist stability criterion. On the basis of the condition, we propose a method for identifying the region that consists of all potential bifurcation boundaries. Finally, robustness of a gene-metabolic oscillator with dynamic uncertainties is investigated by using the proposed method.