Spoke tension and how to true a bike wheel

The evidence here is moderate: real tests, but not many of them. 8 sources, listed at the end.

What the tests say

What tension does. A spoked wheel works because tensioned spokes suspend the rim around the hub¹. Higher tension doesn’t make a wheel stiffer. Damon Rinard’s lateral-load tests found stiffness “does not vary significantly with spoke tension unless a spoke becomes totally slack”², and Wikipedia’s summary also says tension has little influence on stiffness³. What tension does is keep spokes from going slack under load. Roues Artisanales’ stiffness tests show spokes losing tension under heavy lateral loads, which changes how the wheel behaves⁵. Sheldon Brown explains that spokes which repeatedly slacken and re-tension fatigue and break, much like bending a paperclip back and forth. That’s a particular risk on the low-tension side of a dished rear wheel⁴.

How much. Park Tool says most rims suggest 100–120 kgf (980–1,177 N), assuming no pressure in the tire, and advises consulting the rim maker for the right figure¹. For evenness, spokes within ±20% of the wheel’s average are “generally considered” acceptable¹. Park’s Wheel Tension App defaults to ±20% but lets builders choose 15%, 10% or 5%. It balances left-side spokes against each other and right-side spokes against each other, and its worked example averages 60 kgf on the left and 107 kgf on the right (Park doesn’t label the example a rear wheel)⁶.

Drive vs non-drive. On a dished rear wheel (and other asymmetric wheels), the spokes on one side sit at a steeper angle, so the two sides can’t carry equal tension. Park’s guidance is to apply the target tension to the tighter side and let the other side fall where correct dish puts it¹. Sheldon Brown says the same: if the drive side is right and the wheel is correctly dished, the left side “will be quite a bit looser”. Thinner spokes on the low-tension side are less likely to go slack⁴.

Measuring. A tensiometer such as Park’s TM-1 converts spoke deflection to kgf using a table for each spoke diameter and material¹. John Allen’s pitch method gives a rough check. On a typical 700c three-cross wheel, plain-gauge spokes should ring about a G and butted spokes about an A, with no more than a semitone or two of variation. Allen warns against truing by trying to equalize pitch⁷.

How true is true. Park Tool’s targets⁸:

• Lateral (side-to-side): 0.5mm or less, about five sheets of paper
• Radial (up-and-down): 1mm or less

Work lateral first, then radial. Adjust in 1/4 turns, using 1/8 for small errors and up to 1/2 for large ones. Tighten spokes rather than loosening where possible, and adjust radial errors in pairs⁸.

Making it stay true. Spokes twist as their nipples turn (wind-up) and later unwind, letting the wheel drift out of true. Sheldon’s fix is to turn slightly past the target and back off. He also recommends stress-relieving new wheels until they stop pinging and drifting⁴. Park’s version is to ride the wheel and expect some normal pinging at first⁸.

When to stop. Park says that for a rim bent beyond repair by spoke adjustment, truing is unlikely to help except to get the rider home, and the job is best left to a professional mechanic. Park’s rule of thumb: “3 broken spokes = 1 new wheel”⁸.

The numbers

• 100–120 kgf (980–1,177 N): typical rim range. Defer to the rim maker¹
• ±20% of average tension, per side: Park’s acceptable spread. Tighter tolerances (15/10/5%) are optional¹⁶
• Park’s worked example: 60 kgf left vs 107 kgf right⁶
• Lateral true ≤0.5mm; radial ≤1mm⁸
• Adjustment steps: 1/8, 1/4 or 1/2 turn depending on the error⁸
• Pitch method: within a semitone or two (700c three-cross, about G plain-gauge / A butted)⁷
• 3 broken spokes = new wheel (Park’s rule of thumb)⁸

Myths and disagreements

• “Tighter spokes make a stiffer wheel.” Not unless spokes were going slack. Stiffness barely changes with tension²³.
• “Rear-wheel spokes should all be the same tension.” Dish makes the non-drive side lower. Balance each side separately¹⁴⁶.
• “True it by ear.” Pitch is a rough check, not a truing method⁷.
• Disagreement: how even is even enough. Park’s ±20% default is a shop standard. Many builders work to tighter tolerances, and the app allows down to ±5%⁶. No research ties a specific tolerance to wheel life.

Truing a wheel at home

• For a small wobble, use the right-size spoke wrench and quarter turns. Check lateral true against brake pads or a zip tie on the frame, then ride the wheel and re-check⁸.
• If you build or regularly true wheels, get a tension meter and follow the rim maker’s maximum tension¹.
• As a quick check, squeeze pairs of spokes on each side. Compare like with like, since rear non-drive spokes are meant to be looser¹⁶. A spoke clearly looser than its same-side neighbors is worth a closer look. That’s a common shop habit, not a tested method.
• Take it to a pro if the rim is bent beyond what spoke adjustment can fix, as Park advises⁸. Repeated spoke breakage, a cracked rim, a wheel that won’t hold true after stress-relieving, or a carbon rim with specific tension specs are also sensible reasons to see a pro (a suggestion, not from Park).

Sources