What the CWSS Database Actually Holds

The Civil War Soldiers and Sailors System (CWSS) is larger than a name lookup. It indexes about 6.3 million General Index Cards covering roughly 3.5 million soldiers from 44 states and territories, then adds about 4,000 regiment narratives, summaries of 384 battles, around 18,000 African American Navy personnel records, more than 1,200 Medal of Honor citations, and burial records that begin with Poplar Grove National Cemetery at Petersburg. Knowing which of those layers a search result comes from changes how much weight you should give it.

What does the CWSS actually index?

The soldier layer is the General Index Cards created from the Compiled Military Service Records at the National Archives. The National Park Service states the database holds "6.3 million index cards" for roughly 3.5 million soldiers across 44 states and territories, both Union and Confederate. The gap between the card count and the soldier count exists because one man could generate several cards, through multiple regiments, multiple appearances, or spelling variants. For the mechanics of that gap, see Name Spelling Variants on Index Cards.

Does the database include anyone besides soldiers?

Yes, and these layers are easy to miss if you only search a surname. The CWSS lists about 18,000 African American sailors who served in the Union Navy, more than 1,200 recipients of the Medal of Honor, and burial information that, in its first phase, comes from the written records of Poplar Grove National Cemetery at Petersburg National Battlefield. Each of these is a distinct dataset with its own scope, not a filter on the same soldier cards. A sailor's Navy record, a Medal of Honor citation, and a cemetery burial entry can each exist for the same man without any one of them substituting for the others.

Where do the regiment and battle summaries come from?

This is the layer researchers most often mistake for a primary record. The roughly 4,000 regiment narratives are drawn from two published reference works: Frederick Dyer's Compendium of the War of the Rebellion for Union units and Joseph Crute's Units of the Confederate States Army for Confederate units. The 384 battle summaries come from a 1993 report to Congress by the Civil War Sites Advisory Committee. Both are later compiled syntheses, written decades after the war (Dyer began his work in 1903), not transcriptions of a wartime document. That distinction matters enough to deserve its own treatment; see Regimental Histories and Battle Summaries: Reading a Secondary Source.

Why are there more index cards than soldiers who served?

An estimated 3.5 million individuals served, yet the General Index Cards number about 6.3 million. The National Park Service attributes the difference to soldiers who served in more than one regiment, soldiers who served under more than one name, and spelling variations. None of those reasons implies an error in the count; they describe how an index built from repeated appearances naturally outgrows the number of people it describes.

What can the database not tell you?

It cannot tell you, by itself, whether a specific soldier fought in a specific battle. The National Park Service is explicit that linking a soldier to a regiment to a battle "does not indicate with certainty whether a soldier participated in a particular battle, only that the regiment was present." A regiment narrative or battle summary describes the unit's documented movements, not any one man's actions inside them. The compiled card behind a name is itself a later abstraction of a muster roll, created from the 1880s onward, so even the soldier layer is a step removed from the original document. For what a card can and cannot settle on its own, see The Index Card Is Not the Final Answer.

How should this shape a first search?

Before you read a result as a fact about a person, identify which layer produced it.

Result you see	What it actually is	How much it proves about your soldier
A General Index Card	A summary card giving name, unit, and rank in and out, compiled from the 1880s onward	Confirms a name in a unit, not a life story
A regiment narrative	A later compiled summary (Dyer or Crute)	Describes the unit, written decades later
A battle summary	One of 384 entries from a 1993 Congressional report	Confirms the regiment's documented presence, not the man's
A Medal of Honor or Navy record	A separate, purpose-built dataset	Applies only if your soldier matches that specific dataset
A cemetery burial entry	Poplar Grove National Cemetery records (first phase)	Confirms burial, not service details

Treat the first search as a map of where to look next, not a finished answer. The broader the database, the more useful it is to know which drawer a given result came from before you build a conclusion on top of it. For the practical habit of reading one original record against another instead of stopping at the first index hit, Muster Roll vs Compiled Service Record is the place to start.