The honest threshold is a crack longer than a dollar bill — roughly six inches — or any crack that has reached the edge of the glass, and at that point a home repair kit is the wrong tool and a replacement is the right one.
That number isn't arbitrary. The resin kits sold at auto parts stores are built around a specific failure mode: a small, contained break with intact glass around it. Once a crack runs past a few inches, the resin can't be pulled into the full length of the damage before it cures, and the part you didn't fill becomes the part that spreads. The kits still sell, still come with a bridge and a syringe and a curing strip, and still work fine on a quarter-sized star or a bullseye. They just stop working somewhere around the length of a banknote, and most people find that out after they've already spent the money.
Length And Location
Length gets all the attention, but location decides more cases than length does. A crack that terminates in the middle of the windshield is a different animal from one that has reached the perimeter, because the edge is where the glass is bonded to the body and where stress concentrates. A crack that has touched the edge has already relieved some of the tension in the panel, and it will keep traveling along the path of least resistance — usually further along the edge, sometimes inward across the driver's line of sight.
The practical read: measure the crack, then look at where both ends sit. Under about an inch and fully enclosed by intact glass, a kit is reasonable. Between an inch and the length of a dollar bill, it depends on whether the damage is in the driver's primary viewing area and whether it's a single clean line or a branching mess. Past a dollar bill, or touching the edge, or branching into multiple legs, stop. You're not repairing anything at that point; you're decorating a crack that's going to move.
There's also a type question that gets ignored. A chip — bullseye, star, combination — has a defined boundary. A crack has a tip, and the tip is where the glass is actively separating. Resin applied to a crack tip doesn't stop the tip; it fills what's already open. The tip keeps going, often within days, and often when you least want it to: a cold morning, a pothole, a slammed door.
Why Kits Fail
Home kits work by vacuum and pressure. You seal a bridge over the break, pull a vacuum to evacuate air from the fracture, then push resin in and cure it with UV. On a chip, the fracture is a small cavity and the vacuum does its job. On a long crack, the fracture is a narrow channel with a tip that's under load, and the vacuum can't evacuate the whole channel before the resin starts to set. You end up with resin in the first inch and air in the last three.
Temperature matters more than the instructions admit. Resin viscosity changes with temperature, and a cold windshield makes the resin thick enough that it won't wick into a narrow crack at all. A hot windshield cures the resin before it has finished flowing. The window where a home kit actually performs is narrow, and it's narrower still on a long crack, where the resin has further to travel.
The other failure is cosmetic and psychological. A repaired crack doesn't disappear. It leaves a visible line, often with a slight optical distortion, because the resin's refractive index is close to glass but not identical. On a chip in the passenger side, that's invisible in daily use. On a long crack running across the driver's view, it's a permanent distraction that no amount of polishing fixes — and it's the reason many people who try a kit on a long crack end up replacing the windshield anyway, having spent the kit money first.
When To Replace
Replace when the crack is longer than a dollar bill, when either end has reached the edge of the glass, when the damage sits directly in the driver's line of sight, or when the crack has branched. Replace also when the crack is on the inner layer of a laminated windshield — you can sometimes tell because the outer surface feels smooth while the line is clearly visible underneath. That's a delamination-adjacent failure and it isn't a resin job at any length.
There's a safety argument underneath the cosmetic one. Windshields are structural on modern vehicles. They contribute to roof crush resistance and they're the anchor for the passenger-side airbag, which deploys against the glass. A windshield with a long crack has reduced stiffness in the cracked region, and the crack can propagate during a crash event in ways that change how the airbag loads. That's not a reason to panic about a chip, but it is a reason not to drive around for months with a crack running halfway across the glass because a kit seemed cheaper.
Cost is the trade-off people weigh, and it's worth being honest that replacement usually costs several times what a kit costs, and sometimes several times what a professional chip repair costs. If the crack is marginal — right at the edge of repairable — a shop with a proper vacuum cycle and better resin has options a driveway kit doesn't. Getting a shop to look at it costs nothing at most places. Getting it wrong costs a windshield.
Driving On A Crack
A cracked windshield is not automatically an emergency, but it degrades quickly and the degradation is not linear. The first few days after a crack appears are when it moves fastest, because the stress that caused it hasn't fully redistributed. Thermal cycling is the main driver: park in the sun, the glass expands; run the defroster on a cold morning, the inner surface heats before the outer, and the differential loads the crack tip. Every one of those cycles nudges it forward.
Physical shocks do the rest. A pothole, a hard door slam, a speed bump taken too fast — each sends a stress wave through the panel. If you must drive on a long crack, avoid temperature extremes on the glass, don't slam doors, and keep an eye on the ends. If you can see the crack getting longer over a week, it's going to keep going, and the sooner it's replaced the less chance it spreads into a shape that complicates the install.
One thing that does help temporarily: clear tape over the crack, on the outside, to keep water and grit out of the fracture. Water in a crack is bad — it contaminates the surfaces and can freeze and wedge the crack wider. Tape is not a repair. It's a stopgap that buys you a few days until the replacement appointment, and it's worth doing if rain is in the forecast and you can't get in immediately.
Insurance And Timing
Many policies cover windshield replacement with a lower deductible than collision, and some cover chip repair with no deductible at all because insurers would rather pay for a repair than a replacement. That asymmetry is worth checking before you decide, because it can flip the math entirely: if repair is free and replacement is a small deductible, the argument for a home kit disappears.
The timing question is simpler. A crack that's repairable today may not be repairable next week, because the tip moves. If you're on the fence about whether yours is too long, the cost of having a shop look at it is near zero and the cost of waiting is a windshield. Get it assessed while it's still marginal, and if the answer is replace, book it before the crack decides for you.