Abstract A novel nonlinear adaptive prediction approach based on truncated Laguerre memory structure is proposed to effectively suppress NarrowBand Interference (NBI) in Direct Sequence Spread Spectrum (DSSS). The memory depth can be uncoupled with the filter orders. Simulation results show that the proposed nonlinear adaptive prediction scheme outperforms the other exiting nonlinear adaptive prediction methods for the NBI suppression in DSSS, and the orders of the proposed filter can be largely reduced 3~5 times than that of Vijayan and Poor’s filters.