In the oligohaline Alloway Creek watershed of the upper Delaware Bay, invasive Phragmites australis (Common reed; hereafter Phragmites) has been removed in an attempt to restore tidal marshes to pre-invasion form and function. In order to determine the effects of Phragmites on nekton use of intertidal creeks and to evaluate the success of this restoration, intertidal creek nekton assemblages were sampled with weirs from May to November for 7 years (1999-2005) in three marsh types: natural Spartina alterniflora (Smooth cordgrass; hereafter Spartina), sites treated for Phragmites removal (hereafter referred to as Treated), and invasive Phragmites marshes. Replicate intertidal creek collections in all three marsh types consisted primarily of resident nekton and were dominated by a relatively low number of ubiquitous intertidal species. The Treated marsh nekton assemblage was distinguished by greater abundances of most nekton, especially Fundulus heteroclitus (Mummichog). Phragmites had little impact on nekton use of intertidal creeks over this period as evidenced by similar nekton assemblages in the Spartina and Phragmites marshes for most years. Long-term assemblage-level analyses and nekton abundances indicated that the Treated marsh provided enhanced conditions for intertidal creek nekton. The response of intertidal creek nekton suggests that the stage of the restoration may influence the results of comparisons between the marsh types and should be considered when evaluating marsh restorations.