Forest inventory data on the national scale cover a broad range of information. Attributes on forest production still play an important role, however ecologically relevant data on forest structure and species diversity get increasing attention. Typically, forest inventories are sample based and the estimation of ecological relevant information from inventory data is restricted. Indicators as e.g. patch perimeter are usually measured from maps, if those indicators cannot be measured directly - e.g. due to the absence of maps- it might be of interest how they could be estimated from sample based inventory data. Methods to estimate the forest boundary length are presented and applied to data of the second Swiss National Forest Inventory (NFI). The forest area definition of the Swiss NFI is based on fine grain data derived from air photo plots classified to nonforest and forest categories. Both, original photo plot interpretation data as well as the relationship between the classification results; (1) interior forest plots, (2) nonforest plots and (3) forest plots intersecting a forest border are employed. The presented methods allow to estimate forest boundary length figures from successive inventory occasions and thus add a landscape ecological component to already established sample based inventory systems.