This repository contains code for developing an open source method to derive small area estimates of linked (origin-destination) commuting trips by transportation mode from open source Longitudinal Employer-Household Dynamics (LEHD) Origin-Destination Statistics (known as LODES) and American Community Survey (ACS) data in R. This method is fully-customisable and can potentially be applied to the entire US (with a small number of exceptions) for every year from 2002-2019. For demonstration purposes, the code here applies this method to 2015 commuting flows at the tract scale in Cook County, Illinois. Although the data used in this application can be downloaded openly from the US Census Bureau (2015 LODES: https://lehd.ces.census.gov/data/; 2013-2017 5-year ACS: https://data.census.gov/cedsci/), lightly-cleaned version are provided in this repository for convenience. Obtaining tract-to-tract transit travel times from r5r, which is used to determine viable transit trips, also requires an OpenStreetMap (OSM) file for a given region and a valid General Transit Feed Specification (GTFS) file. While these can also be obtained openly online (from https://www.interline.io/osm/extracts/ and http://transitfeeds.com/p/chicago-transit-authority/165, respectively), versions for 11 November, 2021 for Chicago are also included in this repository. The 'LODES-ACS-commuting-flows.R' code file contains the code necessary to run our application, with pre-computed tract-to-tract travel times.