We analyzed 136 time series (covering from 44 to 73 yr) of juvenile cod to estimate the level of direct and delayed density-dependent mortality (DDM) of 11 populations from the Norwegian Skagerrak coast. The parameters were estimated using a modeling approach that explicitly incorporates observation errors, so that we could quantify the density-independent (stochastic) variation in the survival of juvenile cod. Moderate to strong levels of DDM (direct or delayed) were estimated in eight of the 11 populations. Variability in the 0-group (corrected for observation errors) appeared to be large for most of the populations. Substantial stochastic variability in postsettlement survival was also detected in some areas, indicating that stochastic factors are not only important for egg and larval stages, as stated by the match-mismatch hypothesis, but also for juveniles. We show that the variability in these coastal populations is not only regulated as a function of the strength of DDM processes, but also as an interaction between DDM processes and stochastic factors. We finally postulate that local and regional differences in the strengths of the density-dependent and stochastic processes are related to differences in the quantity and quality of the bottom flora coverage, which govern both food availability and shelter for juveniles.