Recently released NCHRP Report 770 offers a menu of tools for estimating non-motorized travel demand. Among the new tools created is an accessibility-based approach that relies on GIS data and methods to create the necessary relationships for travel demand analysis. By layering a digitized rendition of the travel network onto the prevailing land use, it is possible to create measures of accessibility to employment, retail or other activities for each mode at any point in space. When these accessibility measures are linked to trip ends from a travel survey, strong relationships are seen between the modal accessibilities and chosen mode. Equations thus created using disaggregate trip data can then be applied back at the same (individuals) or more aggregate levels of geography (parcels, census blocks) for planning analysis. The original models were developed using data for Arlington County, VA and test-applied with good success to several different community settings in the County. A particularly illustrative example involved application to the town of Shirlington, a popular and rapidly evolving mixed use/walkable center. Application of the model to existing conditions confirmed high levels of walk accessibility; however, the analysis also revealed some critical disconnects in the existing walk network that discouraged walking from a high rise residential neighborhood in the west to the commercial activities in the east. By inserting strategic enhancements to the network and recalculating the accessibilities, the model revealed a substantial increase in pedestrian demand between these two locations, demonstrating its practical utility as a planning tool; the model could as easily been used to examine the value of siting more non-residential land uses in the vicinity of the residential areas. The model is currently being used as the principal tool in the analysis of a large multi-modal corridor, which is also expanding its capability for transit analysis (reflecting the importance of non-motorized access and accessibility at both trip ends). Exciting details of this new work will be shared if the results can be made public at that time (as is expected).