Runnels, a climate adaptation technique that drains surface water to restore marsh vegetation and habitat, are increasingly being used to prevent the formation of shallow water impoundments or pannes in salt marshes that result in the loss of important ecosystem services. However, we know little about the effect of runnels on salt marsh biogeochemistry. This study measured how sediment characteristics and rates of nitrogen cycle processes were altered by impounded water and vegetation loss, and whether runnels can restore these marsh attributes to reference conditions. Impounded areas were 52 ± 4% less vegetated than nearby intact marsh, with 11 ± 2% less organic matter and 24 ± 5% higher bulk density. Additionally, impoundments removed 32 ± 32 µmol N m−2 d−1 less than reference marsh areas via denitrification. At six of the 11 runneled sites, vegetation percent cover increased by 40 ± 5%, accompanied by a 7 ± 3% recovery of organic matter and a 9 ± 6% reduction of bulk density. At sites where vegetation recovered to within 70% of reference plots at a site, runneled plots removed 97 ± 31 µmol more N m−2 d−1 than impoundments, which was also 82 ± 31 µmol more N m−2 d−1 than reference areas. The driver of recovery is related to initial site conditions, including higher redox potentials and lower porewater salinities, compared with sites where revegetation was unsuccessful. The extent of runnel effectiveness and the recovery of vegetation, sediment characteristics, and nitrogen cycle processes was variable among runneled marshes, and the effectiveness of runnels may depend on initial site-specific characteristics and degree of initial degradation.