Estimates of distributions of natal dispersal distances and juvenile recruitment rates in open populations are strongly influenced by the extent and shape of the areas sampled. Techniques to improve biased dispersal and survival estimates include area-ratio methods based on weighting observations by sampling effort, the extent and shape of the area sampled, and the amount and distribution of preferred habitat surrounding the area sampled. We partitioned territories within the boundaries of a large, almost geographically closed, population of individually marked Red-cockaded Woodpeckers (Picoides borealis) and estimated dispersal and survival parameters from hypothetical smaller study areas (sampling areas) of varying sizes and shapes in order to examine whether an area-ratio method provides accurate or improved estimates of juvenile dispersal distance and survival. Non-aggregated sampling areas resulted in the detection of fewer dispersal events, but because of their large spatial extent, produced unbiased dispersal estimates. The use of aggregated sampling areas (circular or linear) resulted in the detection of higher numbers of dispersal events, but produced biased dispersal estimates that were generally improved by the area-ratio method. Area-ratio corrections usually provided better estimates of median dispersal distance than uncorrected estimates. Survival to breeding was usually underestimated and often not improved by the area-ratio method, regardless of extent and shape of the sampling area. Estimates of juvenile survival to breeding were improved by assuming that rates of emigration were equivalent to immigration, and correcting survival estimates accordingly. Small, local studies should use an area-ratio method to improve their estimates of median dispersal distance. Because the correction method estimates relative, but not absolute, numbers of individuals dispersing across distance categories, the area-ratio method should not be used for estimating survival. Non-aggregated sampling areas may be an effective design to increase spatial extent (and thus decrease bias) without proportionately increasing the amount of habitat sampled.