In the process of oilfield development, as water injection continues, the increase in oilfield water content has a significant impact on the economic benefits and development strategies of the reservoir. Especially in Block S, the oil field has entered a stage of high extraction difficulty and low oil production, at which point the identification of water flooded layers becomes particularly important. The traditional method for identifying water flooded layers faces challenges such as high dynamic complexity and difficulty in classification, especially in the regression fitting of reservoir parameters such as porosity (PORE), permeability (K), saturation (SW), etc. This article proposes a water flooded layer identification method based on quantum neural network (QNN) to address the small sample space problem using valuable core data obtained from exploration and development in S block. This method first constructs a quantum neural network model that can handle multi-dimensional discrete sequence samples. The model uses quantum neurons as hidden layers and ordinary neurons as output layers, and designs a learning algorithm based on gradient descent. By selecting nine features that describe the level of flooding, the network is trained by directly constructing training samples using the discrete values of these features. The trained network can be effectively applied to water flooded layer recognition tasks in similar regions. Taking 123 formation samples from 11 wells in Block S as an example, the water flooded layer recognition method proposed in this paper achieved a recognition rate of 86%, demonstrating the good adaptability and practicality of quantum neural networks in automatic recognition of water flooded layers. This study not only improves the accuracy of identifying water flooded layers, but also provides strong technical support for oilfield development decision-making.