Glass is generally considered class evidence, meaning it can be grouped with other samples that share similar properties but cannot typically be traced back to one unique source. The exception is when broken glass fragments can be physically pieced back together like a jigsaw puzzle, which elevates those specific pieces to individual evidence. Understanding this distinction matters because it directly affects how much weight glass carries in linking a suspect to a crime scene.
Class vs. Individual Evidence
Forensic evidence falls into two broad categories. Individual evidence can be linked to a single, unique source with near certainty. DNA is the classic example: testing thirteen specific locations on a DNA molecule produces a profile so specific it identifies one person. Class evidence, by contrast, narrows the field to a group of possible sources without pinpointing one. A white cotton fiber found at a scene could have come from millions of identical white cotton shirts.
Glass lands in the class category for a straightforward reason. Modern glass manufacturing, particularly the float process used for windows and vehicles, uses tight quality control that produces enormous quantities of nearly identical material. Two sheets of glass made at the same plant in the same week can be chemically indistinguishable from each other. Unlike DNA, where every person’s profile is different, glass from a single production batch shares the same basic recipe. Forensic scientists have not been able to build population databases for glass that are anywhere near as useful as those for DNA, precisely because so many samples look alike under testing.
What Makes Glass Useful as Class Evidence
Even though glass is class evidence, forensic labs can measure its properties with impressive precision, and those measurements are powerful at ruling sources out. When a questioned glass fragment does match a known source across multiple tests, that narrows the possibilities considerably.
The primary measurement is refractive index, which describes how much a piece of glass bends light. The FBI’s standard method uses a system called GRIM 3, which immerses a tiny glass fragment in liquid and slowly changes the temperature until the glass and liquid bend light identically. This instrument is precise to the fifth decimal place, meaning it can detect extremely small differences between two samples. If a fragment from a suspect’s clothing and a fragment from a broken window have different refractive indices, they did not come from the same source. If they match, they could have come from the same source, but so could glass from the same manufacturing batch.
Density is another measurable property. Labs use methods like suspending fragments in carefully calibrated liquids inside a temperature-controlled bath, then determining the exact density at which the glass neither sinks nor floats. Beyond these physical measurements, elemental analysis identifies the chemical fingerprint of a glass sample by measuring concentrations of seventeen or more major and trace elements. Studies using laser-based sampling paired with mass spectrometry have shown that glass from the same production run is highly homogeneous, typically varying less than 5% across samples. But glass made at different plants, or even at the same plant weeks apart, often shows measurable differences. This batch-to-batch variation is what gives forensic glass comparison its discriminating power.
When Glass Becomes Individual Evidence
The one scenario where glass crosses the line into individual evidence is the physical fracture match. If a fragment recovered from a suspect’s shoe fits perfectly into a broken window like a puzzle piece, that fragment can only have come from that specific window. The irregular, complex path a crack takes through glass is essentially unrepeatable, making the fit unique.
Forensic examiners evaluate fracture matches using comparative microscopy, aligning the jagged edges of two pieces and checking that the surface features correspond at both the visible and microscopic level. This process relies heavily on examiner experience and judgment. A 2009 report from the National Academy of Sciences noted that fracture matching, like many pattern-comparison disciplines, lacks a strong statistical foundation for estimating error rates. Newer research is developing quantitative methods that map fracture surface topography and apply statistical learning to calculate match probabilities, but in current practice, the examiner’s visual and tactile assessment remains the standard approach.
For a physical match to work, the recovered fragments need to be large enough and intact enough to align. That is why evidence collection protocols emphasize preventing further breakage. Labs recommend securing glass in heavy cardboard packaging or boxes with cushioning material, and they specifically warn against using tape, which can damage fragment edges.
Fracture Patterns and What They Reveal
Even when fragments are too small for a physical match, the way glass breaks provides investigative information. When a force strikes a pane, radial cracks shoot outward from the impact point in a star-shaped pattern, while concentric cracks form circular rings around it. Along the edges of radial cracks, tiny curved ridges called hackle marks appear. These ridges follow a predictable rule: on radial fractures, the ridges are at right angles to the side of the glass opposite the impact. This lets examiners determine which direction the force came from, which matters in cases involving gunshots or break-ins. High-velocity impacts also leave a characteristic cone-shaped crater on the exit side of the glass.
These fracture patterns are class-level information. They tell investigators about the type and direction of force but do not connect the glass to a unique source.
How Forensic Labs Report Glass Comparisons
Because glass is class evidence, labs cannot say a fragment “came from” a particular window the way they might link DNA to a person. Instead, the standard language is that the questioned sample and the known source are “indistinguishable” in their measured properties. The FBI protocol spells this out: when the average refractive index of questioned fragments falls within the range of values measured from a known source, the two are reported as indistinguishable.
To add more meaning to that conclusion, some forensic scientists use likelihood ratios. This statistical approach compares two scenarios: how probable the evidence would be if the glass came from the same source versus how probable it would be if it came from a different source. A study using glass of known manufacturing history found that likelihood ratios ranged from low values when samples came from different plants to very high values when samples were made on the same day at the same facility. This gives courts a more nuanced picture than a simple “match” or “no match.”
Transfer and Persistence on Clothing
A common question in casework is whether glass fragments on a suspect’s clothing actually link them to a breaking event or could have arrived some other way. Research on indirect transfer, where someone touches broken glass and then touches their jacket, shows that fragments move readily. In seven out of nine tests, ten or more fragments were still on a jacket a full 60 minutes after contact, and in three of those tests, more than twenty fragments were recovered. Persistence drops as the person moves around, with fragments gradually falling off during activity like walking, but the initial transfer is robust enough that even indirect contact can deposit a meaningful number of fragments.
This matters for evidence interpretation. Finding glass on clothing is significant, but it does not by itself prove someone broke a window. The fragments need to be compared against a known source using refractive index, density, and elemental analysis. And even a strong match only places the glass in the same class, not at the same unique origin, unless a physical fracture match exists.