<p>This paper focuses on utilizing adaptive sliding mode control (ASMC) to achieve predefined-time synchronization of fuzzy competitive neural networks with proportional delay. Initially, a novel predefined-time stability theorem is proposed according to a fixed-time stable theorem and related inequalities. Unlike previous predefined-time stability results, the proposed criterion is more comprehensive and provides better conservatism with adjustable parameters, making it highly suitable in practical application. Secondly, ASMC scheme is implemented by formulating double-layer controllers and sliding mode surfaces to stabilize the error state in the drive–response system and approach zero on the sliding manifold within a predefined time. Compared to existing sliding mode control technology, the parameters of this controller can be defined as arbitrary constant in advance, which makes the system state of the neural network more superior in estimating the convergence time, and it is more convenient to adjust the controller parameters. Ultimately, we validate the effectiveness of the theorems in this study by offering two examples and application.</p>