This paper deals with the problem of ℒ2‐ℒ∞ model reduction for continuous‐time nonlinear uncertain systems. The approach of the construction of a reduced‐order model is presented for high‐order nonlinear uncertain systems described by the T‐S fuzzy systems, which not only approximates the original high‐order system well with an ℒ2‐ℒ∞ error performance level γ but also translates it into a linear lower‐dimensional system. Then, the model approximation is converted into a convex optimization problem by using a linearization procedure. Finally, a numerical example is presented to show the effectiveness of the proposed method.