<p>Lateral weirs, also known as side weirs, are significant hydraulic structures used to regulate and divert flow from primary channels. These structures are essential in managing water distribution in irrigation networks, canal systems, and urban drainage infrastructures. Previous studies have primarily focused on conventional trapezoidal, triangular, and arced labyrinth configurations, limited research is available on rectangular labyrinth side weirs; hence, this study investigates the hydraulic behaviour of a two-cycle rectangular labyrinth side weir. Accurately predicting the discharge coefficient is key to determining the outflow characteristics of such structures. A detailed experimental investigation was conducted in the laboratory to evaluate the discharge behaviour. The resulting data were used to develop predictive models employing non-linear regression and artificial neural network (ANN) techniques. Comparative analysis using statistical performance indicators reveals that the ANN model exhibits strong predictive performance, as indicated by a highcoefficient of determination (R²) value of 0.9586 and root mean square error (<i>RMSE</i>) of 0.0186. Sensitivity analysis further identifies tamong the parameters considered, the Froude number (<i>Fr</i>) had the greatest impact on predicting the discharge coefficient.<i>.</i> This study presents comprehensive experimental datasets and ANN-based prediction frameworks for the discharge coefficient of a two-cycle rectangular labyrinth side weir, a geometry that has received limited attention in previous research. The results extend the applicability of data-driven modelling technique to rectangular labyrinth configurations and provide a reliable basis for their hydraulic design.</p>