Simultaneous measurements of vertical accretion from artificial soil marker horizons and soil elevation change from sedimentation-erosion table (SET) plots were used to evaluate the processes related to soil building in range, basin, and overwash mangrove forests located in a low-energy lagoon which recieves minor inputs of terregenous sediments. Vertical accretion measures reflect the contribution of surficial sedimentation (sediment deposition and surface root growth). Measures of elevation change reflect not only the contributions of vertical accretion but also those of subsurface processes such as compaction, decomposition and shrink-swell. The two measures were used to calculate amounts of shallow subsidence (accretion minus elevation change) in each mangrove forest. The three forest types represent different accretionary envrionments. The basin forest was located behind a natural berm. Hydroperiod here was controlled primarily by rainfall rather than tidal exchange, although the basin flooded during extreme tidal events. Soil accretion here occurred primarily by autochthonous organic matter inputs, and elevation was controlled by accretion and shrink-swell of the substrate apparently related to cycles of flooding-drying and/or root growth-decomposition. This hydrologically-restricted forest did not experience an accretion or elevation deficit relative to sea-level rise. The tidally dominated fringe and overwash island forests accreted through mineral sediment inputs bound in place by plant roots. Filamentous turf algae played an important role in stabilizing loose muds in the fringe forest where erosion was prevalent. Elevation in these high-energy environments was controlled not only by accretion but also by erosion and/or shallow subsidence. The rate of shallow subsidence was consistently 3-4 mm y-1 in the fringe and overwash island forests but was negligible in the basin forest. Hence, the vertical development of mangrove soils was influenced by both surface and subsurface processes and the procces controlling soil elevation differed among forest types. The mangrove ecosystem at Rookery Bay has remained stable as sea level has risen during the past 70 years. Yet, lead-210 accretion data suggest a substantial accretion deficit has occurred in the past century (accretion was 10-20 cm < sea-level rise from 1930 to 1990) in the fringe and island forests at Rookery Bay. In contrast, our measures of elevation change mostly equalled the estimates of sea-level rise and shallow subsidence. These data suggest that (1) vertical accretion in this system is driven by local sea-level rise and shallow subsidence, and (2) the mangrove forests are mostly keeping pace with sea-level rise. Thus, the vulnerability of this mangrove ecosystem to sea-level rise is best described in terms of an elevation deficit (elevation change minus sea-level rise) based on annual measures rather than an accretion deficit (accretion minus sea-level rise) based on decadal measures.