Determining how post-settlement processes modify patterns of settlement is vital in understanding the spatial and temporal patterns of recruitment variability of species with open populations. Generally, either single components of post-settlement loss (mortality or emigration) are examined at a time, or else the total loss is examined without discrimination of mortality and emigration components. The role of mortality in the loss of early juvenile blue crabs, Callinectes sapidus, has been addressed in a few studies; however, the relative contribution of emigration has received little attention. We conducted mark-recapture experiments to examine the relative contribution of mortality and emigration to total loss rates of early juvenile blue crabs from seagrass habitats. Loss was partitioned into emigration and mortality components using a modified version of Jackson's (1939) square-within-a-square method. The field experiments assessed the effects of two size classes of early instars (J1-J2, J3-J5), two densities of juveniles (low: 16 m-2, high: 64 m-2), and time of day (day, night) on loss rates. In general, total loss rates of experimental juveniles and colonization rates by unmarked juveniles were extremely high (range = 10-57 crabs m-2/6 h and 17-51 crabs m-2/6 h, for loss and colonization, respectively). Total loss rates were higher at night than during the day, suggesting that juveniles (or potentially their predators) exhibit increased nocturnal activity. While colonization rates did not differ by time of day, J3-J5 juveniles demonstrated higher rates of colonization than J1-J2 crabs. Overall, there was high variability in both mortality and emigration, particularly for emigration. Average probabilities of mortality across all treatment combinations ranged from 0.25-0.67/6 h, while probabilities of emigration ranged from 0.29-0.72/6 h. Although mean mortality rates were greater than emigration rates in most treatments, the proportion of experimental trials in which crab loss from seagrass due to mortality was greater than losses due to emigration was not significantly different from 50%. Thus, mortality and emigration appear to contribute equally to juvenile loss in seagrass habitats. The difference in magnitude (absolute amount of loss) between mean emigration and mean mortality varied between size classes, such that differences between emigration and mortality were relatively small for J1-J2 crabs, but much larger for J3-J5 crabs. Further, mortality rates were density-dependent for J3-J5 juvenile stages but not for J1-J2 crabs, whereas emigration was inversely density-dependent among J3-J5 stages but not for J1-J2 instars. The co-dependency of mortality and emigration suggests that the loss term (emigration or mortality) which has the relatively stronger contribution to total loss may dictate the patterns of loss under different conditions. For older juveniles (J3-J5), emigration may only have a large impact on juvenile loss where densities are low, since the contribution of mortality appears to be much greater than emigration at high densities. The size-specific pattern of density-dependent mortality supports the notion of an ontogenetic habitat shift by early juvenile blue crabs from seagrass to unvegetated habitats, since larger individuals may experience increased mortality at high densities within seagrass beds. Qualitative comparisons between this study and a concurrent study of planktonic emigration of J1-J5 blue crabs (Blackmon and Eggleston, 2001) suggests that benthic emigration among J1-J2 blue crabs was greater than planktonic emigration; for J3-J5 stages benthic and planktonic emigration were nearly equal. This study demonstrates the potentially large role of emigration in recruitment processes and patterns of early juvenile blue crabs, and illustrates how juvenile size, juvenile density, and time of day can affect mortality and emigration rates as well as total loss and colonization. The components of po