<p>This paper considers the bifurcation problem of a fractional-order four neuron recurrent neural network with double delays. Firstly, by analyzing the relevant characteristic equations, the stability of the system and the Hopf bifurcation problem are studied. The obtained results imply that the dynamics of delayed of derivative order of fractions neural networks are also significantly affected by applications with different delays. Secondly, the influence of fractional orders on bifurcation points is illustrated through drawing comparison, and the unstable amplitudes of the system in derivative order of fractions and integerorder states were also compared. Finally, the effectiveness of the theoretical results is tested by means of numerical examples.</p>