An extended state observer-based multiple description coding data-driven adaptive sliding mode control (ESO-MDC-DDASMC) method is proposed for a class of discrete-time nonlinear nonaffine networked control systems (NCSs). First, a dynamic linearization technology is employed to transform the nonlinear NCSs into a data model with a linear parameter term and a nonlinear residual term. Then, to improve data transmission reliability, an MDC protocol is introduced in the sensor-to-controller channel. The MDC encodes the source signal into two descriptions, which are transmitted through independent channels, and introduces two independent Bernoulli random variables to characterize the data dropout phenomenon of each channel. On this basis, the ESO-MDC-DDASMC method is proposed with both a parameter updating algorithm and an extended state observer. The former is to estimate the unknown parameters and the latter is to estimate the uncertainties and disturbances. Then, the boundedness of the tracking error is proved through the contraction mapping principle. Finally, the effectiveness and applicability of the proposed method are demonstrated by the numerical simulation and practical examples.