Figuring Out What an Old Key Actually Is
I got a bin of thirty-something keys from an estate sale last year. Some were brass skeleton keys, some look like they came from a filing cabinet in 1972, and one was a warded key that probably opened a 19th-century door in rural Kentucky. Sorting through them took longer than I expected because most of the online guides gloss over the hard parts. They show you the textbook shapes but don't explain what happens when the bitting is worn down to a rounded puddle or when the key has been re-cut four times over its life. What I've learned sitting at my workbench with a magnifying lamp and a digital caliper is that identification starts with the lock mechanism type, not the key shape. Key shape is misleading because locksmiths frequently recut keys onto different blank profiles. You can have a key that looks like a standard Warded key but actually fits a tubular lock because someone filed the edge to make it work. The mechanism is the anchor. Everything else is secondary.
Getting Started With an Antique Key Identification Guide Approach
Start by measuring the key blade length from the shoulder to the tip. Then measure the blade width at its widest point. These two numbers get you into the ballpark of which manufacturer or era you're dealing with. A typical Victorian warded key from the 1880s runs about 3.5 inches long with a blade width around 0.4 inches. A yale style pin tumbler key from the 1920s is closer to 2.75 inches with a narrower blade around 0.3 inches. But these are rough ranges. I've seen regional variations where a Pennsylvania locksmith made keys three-eighths of an inch longer than the same design from Ohio, and the locks weren't interchangeable. Look at the bow shape next. Bow design is one of the most underused identification markers. Manufacturers had distinct bow styles, and some were trademarked. A Brown & Sharpe bow has a very specific rounded profile with machined grooves that are hard to replicate. A Silsbee bow tends to be more angular with a pronounced thumb rest area. When I'm trying to date a key, I compare the bow to reference photos first before touching the blade. The bow survives wear better than the cuts because it doesn't go inside the lock. I've found bows on keys that are decades older than what the bitting would suggest, which means the key was recut onto an older blank. The bitting pattern itself is where things get complicated. On a warded key, the cuts are designed to clear internal wards or obstacles inside the lock. Each cut depth corresponds to a clearance requirement, not a pin height. Measuring warded key bitting with a caliper gives you approximate depths, but the real test is whether the key clears the wards when you hold it up to a reference image of the lock interior. Without seeing the actual lock, you're guessing. I once spent two weeks trying to identify a key with what looked like standard warded cut spacing. It turned out to be a Chubb detection key, which uses a different principle entirely. The cuts look similar but the lock internals are completely different. I identified it by noticing the double-slotted cut near the bow, which is characteristic of Chubb detectors. That detail was barely visible without good lighting.
The Practical Stuff Most Guides Skip
Patent dates stamped on the bow are useful but limited. A date of 1865 on a bow tells you the design is at least that old, not that the key is from that year. Keys were produced for decades after patent expiration. I found a key with an 1872 patent date and a 1898 manufacturing mark from the same company. The patent date and production date were twenty-six years apart. Always check for manufacturer marks near the bow. Companies like Yale, American, and Dotzauer put their logos on the bow or the blade shoulder. These marks are often tiny and worn smooth. A soft brass stylus and good side lighting help reveal them without damaging the surface. Material composition matters for dating. Brass keys became common after the 1840s when efficient brass production methods were developed. Before that, keys were often iron or steel, sometimes with brass ferrules. Iron keys from the early 1800s are usually heavier than they look and show different patina patterns than brass. If a key feels disproportionately light for its size, it might be a later alloy or a reproduction. I've seen modern reproductions of Georgian keys that use aluminum-brass alloys which have a distinctly different color and weight. The reproduction keys also tend to have machine-finished surfaces that look too uniform compared to hand-fitted originals. Wear patterns tell you how a key was used. A genuine antique key shows consistent wear on the bitting faces where it contacted the wards or pins. The wear should follow the direction of insertion and rotation. If the wear patterns look random or show cross-hatching, the key was probably tested on multiple locks or was a later copy. I had a customer bring in what he thought was an original 1700s key. The bitting wear was too aggressive and covered the entire cutting face evenly. Original keys from that period show wear concentrated on specific edges where the metal contacted the lock obstacles. This key had been copied and re-cut so many times that the original profile was gone. I could see the original cuts underneath the later work if I angled the light properly. The underlying pattern was a simple warded design, not the complex later cuts.
Get the Full Details
Comparing keys to lock samples is the most reliable identification method. If you can find the original lock or a matching example, test the key carefully. Do not force it. If the key does not slide in smoothly with light pressure, it is not the right match. I keep a collection of lock cylinders from different eras specifically for this purpose. A 19th-century warded lock, an early 20th-century pin tumbler, a tubular lock from a 1950s appliance. Testing against known locks takes maybe ten minutes per key and eliminates most of the guesswork. The alternative is measuring and comparing to photographs, which works about sixty percent of the time depending on the key condition.
When Identification Fails Completely
Sometimes a key is just unidentifiable. Keys that have been heavily re-cut, filed, or repaired lose their original characteristics. A key that has been cut five or six times to fit different locks may have no remaining original profile. In these cases, you can sometimes still determine the lock type by examining the bow and blade dimensions, but the specific manufacturer or model is usually lost. I had a key last year that was clearly a pin tumbler key based on the blade width and cut pattern spacing. It was cut so deeply that the remaining material between cuts was less than a millimeter in places. I could not determine if it was a residential, commercial, or automotive key. The best I could do was estimate it was a single cylinder yale style key from roughly 1930 to 1950 based on the overall proportions. That is about as precise as it gets with a key in that condition. Reproductions and keys complicate everything. The market for antiques means there are a lot of modern keys stamped with old patent dates. These are usually obvious if you know what to look for. Modern reproductions use different metallurgy, have sharper machine marks, and often feature patent dates that do not align with the actual design era. An 1880s-style warded key with a 1962 patent date stamped on the bow is almost certainly a reproduction. The patent system did not work that way in the 19th century. I check patent databases when I find dates I am unsure about. A quick search usually confirms whether a date makes sense for the key design. For keys that resist identification, the practical alternative is having a locksmith cut a working copy from the original if it still fits any lock. Even if you cannot identify the key, a skilled locksmith can often make it functional by testing against compatible blanks and adjusting the bitting. This does not solve the identification problem but it solves the usability problem. I recommend this approach when the key is in your possession and you need it to work rather than when you are cataloging a collection for historical documentation. The two goals require different approaches.
