How do I design a custom ring that matches an existing wedding band?
Marco came in last February with a platinum wedding band his wife had worn for twelve years. Simple 3mm flat court profile, slight satin finish on top,...
Marco came in last February with a platinum wedding band his wife had worn for twelve years. Simple 3mm flat court profile, slight satin finish on top, polished inside. He wanted an engagement ring that sat flush against it. No gap allowed.
This is one of the most common-and most specific-requests I get across the bench. The good news is it's almost always solvable. The bad news is most jewelers won't ask the two questions that actually matter before they start designing.
The two questions nobody asks
First: What's the band profile? Grab a piece of paper and trace the cross-section of the existing wedding band. Is it flat court (flat outside and in)? Comfort-fit (domed outside, flat inside)? Court (fully domed like a D-shaped section)? Beveled? Knife-edge? The answer determines whether a second ring can sit flush at all without a spacer.
Second: Where does the band sit on the finger? That sounds obvious until you realize that a 4mm wide band behaves differently on a size 5 finger than on a size 9. The curvature changes. Matching that curve is what makes the pair feel like they belong together.
The practical approach: three paths
Path one: build the new ring to nest
If the existing band has a flat or slightly domed profile, you have the most options. A cathedral setting on the new ring-those arches that rise from the shank to the head-creates clearance for a flush fit. The band from the new ring can be shaped with a slight concave curve on the inside edge, so it wraps around the existing band's profile rather than fighting it.
I did this for a client named Priya last summer. Her grandmother's band was a 2.8mm court profile in 18k yellow. We made the new engagement ring with a matching court profile but added a slight radius to the inside mating surface-about 0.3mm deeper at the bottom of the shank than at the top. Lip service but critical. The two rings now sit together like they were cast as a set.
Path two: modify the new ring with a notch or contour
If the existing band has a complex profile-knife-edge, or a high-domed court, or anything with decorative elements that extend beyond the simple curve-a notch or contour cut into the new ring's shank is the cleanest solution. The jeweler removes a small channel from the inside edge of the new ring where the existing band will sit. Done well, it's invisible when worn. Done poorly, it looks like a machining error.
I'll be honest: this is one of those jobs where the skill of the bench jeweler matters more than the design. A notch has to be cut with a jeweler's saw, filed to exact fit, then refinished so the patina matches the rest of the shank. That's hand work. Four to six hours if it's right. An hour if someone's rushing.
Path three: a curved or chevron shank
If the existing band is so distinctive-a tension-set center, or a very wide flat band, or a ring with stones set all the way around-that no flush fit is possible, the answer is a curved or chevron-shaped shank on the new ring. The curve wraps around the existing band's profile. The new ring can be designed with a 3-4mm gap that the existing band slides into.
This path works. It's also the one most clients don't love when they see the sketch, because the new ring doesn't stand alone as well. The curve looks odd on its own. You have to be okay with that trade-off: a two-ring set that works together versus a ring that works alone.
Shape matters more than size
Most clients walk in thinking about metal type and stone size. I start thinking about curvature. For a flush fit, the inside diameter of the new ring's shank has to match the outside curvature of the existing band's outer edge. That's not a matter of ring size alone-it's about the specific shape of the band's cross-section.
I measure it with calipers and transfer that curve to paper or CAD. If the bands are different metals-say, platinum and 18k yellow-the design still works because the curvature, not the color, is what makes the fit.
What I won't do
I will not promise a flush fit without seeing the actual band. I've had clients send photos and sketches. Those tell me maybe 70% of what I need. The other 30% is in the hand, feeling how the metal tapers, whether there's any cant or twist in the existing profile, and whether the ring has been resized before (which changes its geometry every time).
I also won't build a ring that requires the existing wedding band to be altered. The existing band is the anchor. You don't change the anchor. You build the new ring to meet it.
The timeline you should expect
A proper matching job takes two appointments. The first: I see both rings, or at least the existing one and full measurements of the finger. The second: I show you a wax or resin model of the new ring alongside the actual band, test the fit, and make adjustments before casting. That's six to ten weeks total, depending on the stone. Anyone promising two to three weeks for a truly flush fit is cutting corners-probably skipping the model stage and hoping the CAD translates perfectly. It often doesn't.
About $180 to $350 in extra labor on top of the ring's base cost, depending on whether we're talking a simple notch or a full cathedral. That's the honest range.
Marco's ring came out well. We did a cathedral solitaire with an old European-cut center-about 1.15 carats, J, a little warm in color, which worked with the 18k yellow that matched his wife's band. The two rings sat together with no perceptible gap. He proposed on a Tuesday. She said yes. Then she asked if the rings felt as good together as they looked, and Marco texted me that night: "They do."