We compare two successional models as guides for restoring native riparian understory species along a 160-km stretch of the Sacramento River in California. In 2001 and 2007, we surveyed cover, frequency, and richness of native and exotic understory species in 15 sites planted (1989–1996) with overstory species to determine whether native understory species colonized naturally (passive relay floristics model). In 2007, we surveyed 20 additional sites (planted 1997–2003) in 14 of which understory species were planted (initial floristics model) to evaluate whether planting accelerated community recovery. We surveyed 10 remnant forests as references for successional trajectories. Mean cover and frequency of natives changed little over time in sites where they were not planted initially; increases in native cover in a few sites were primarily due to a single common species (Galium aparine). Species composition shifted from light-demanding to shade-adapted species, both exotic and native, in response to a doubling of overstory cover. Sites with high intensity understory plantings had greater cover and frequency of native understory species than unplanted sites, but were still low relative to reference forests. Light-demanding natives (e.g., Artemisia douglasiana, Rubus ursinus, and grasses) established in sites where they were planted; however, a shade-adapted species (Carex barbarae) did not survive well. Our research indicates that the passive relay floristics and the initial floristic composition approaches serve to restore a few common native understory species, but that planting species as site conditions become appropriate (active relay floristics model) will be needed to restore entire native understory communities.