In order to create a dataset containing a complete picture of the life experiences of the recruits in the sample, archival records were collected for each recruit, and entered into machine-readable datasets. The records were collected from a variety of sources, broken down into three main groups: military records, Surgeons' Certificates, and census records. The military records consist of archival materials on the recruits' experiences during the war, such as duty rosters, enlistment information, and military hospital-stay data. Also included in the Military, Pension, and Medical Records are materials in each recruit's pension file.

Each time a recruit applied for a pension after the war, either a new pension or a change in his previous pension, he had to undergo a rigorous examination by a board of government-certified doctors. The results of these examinations are recorded in the Surgeons' Certificates, the second major source of data for the Union Army project. The Surgeons' Certificates contain a richly detailed record of each recruit's health from after the war until his death.

Finally, we searched for the records of each recruit in our sample in the U.S. Federal Censuses of 1850, 1860, 1900, and 1910. These censuses contain information about the households which the recruits were living in, and their socio-economic status over time. Data is collected not just about the recruit, but also about the other members of the recruit's household. Using the Census Records , a detailed picture of family structure and socio-economic conditions can be drawn for the members of the sample.

Throughout the separate datasets of the Union Army sample, each recruit can be identified by his unique 10-digit identification number, stored in the variable recidnum.