Kingpin inclination, also called steering axis inclination or SAI, is the inward tilt of the steering axis when you look at the car from the front. Two names, one angle. It is built into the steering knuckle at the factory, there is no bolt to turn to change it, and on most production vehicles it sits somewhere between 6 and 8 degrees.
That last part is why the question in the title gets asked so often. A gauge reads 12 or 15 degrees on a truck that drives straight down the highway, the owner looks up the spec, and the gauge looks broken. It almost never is. This article covers what the angle is, how to read it with a gauge and a set of turn plates, the four things that push a reading off, and the one piece of arithmetic that tells a bent knuckle from a bad measurement.
If camber, caster, and toe are still blurry, read camber vs. caster vs. toe first. This one builds on it.
What KPI and SAI actually mean
Picture a line through the upper and lower ball joints, or through the kingpin on a solid axle. That line is the steering axis, and the wheel pivots around it when you steer. Seen from the front of the car, it is not vertical. The top leans in toward the engine. The angle between that lean and true vertical is kingpin inclination.
It is there on purpose. The inward lean makes the car want to lift slightly as the wheels turn, so the steering tries to return to center on its own. It also pulls the point where the axis meets the ground in toward the tire's contact patch, which keeps bumps and braking from wrenching the wheel.
Two things about it matter for anyone with a gauge in hand.
The angle is fixed. Camber, caster, and toe all have an adjustment somewhere. KPI does not. It is whatever the knuckle was forged to, and the only things that change it are a bent part, a worn kingpin, or a replacement knuckle that was never right for the vehicle.
The gauge can read a wider range than any car uses. The KPI vial on the AutoSolo camber, caster, and KPI gauge runs 0 to 14 degrees in half-degree steps, which is the industry range. Most vehicles live in the middle of it. A reading at the far end of the vial means the setup needs a second look before any parts get ordered.
KPI is not caster, even though they sound alike
Both describe a tilt of the same steering axis, which is how the two get mixed up.
Caster is the tilt seen from the side of the car, with the top of the axis leaning toward the rear. KPI is the tilt seen from the front, with the top leaning toward the center of the car. Same line, two different views.
A real number helps. Ford listed the 1966 Mustang V-8 at 6 degrees 52 minutes of kingpin inclination, which is about 6.87 degrees. That is a typical figure for a passenger car of the period, and it is a useful sanity check when a reading comes in at double that.
The practical difference is what you can do about each one. Caster on many cars can be shimmed or adjusted. KPI cannot. So a caster reading that is off sends you to the adjusters. A KPI reading that is off, once the measurement is proven good, sends you looking for a damaged or wrong part.
How to measure KPI at home
The method is called out, zero, in. It takes the gauge, a set of turn plates, and about five minutes a side once the car is level.
- Set camber to the factory spec first. KPI is read relative to camber, so any camber error lands in the KPI number. Measure camber, adjust it to spec, then move on.
- Turn the wheel 20 degrees out and zero the gauge. Out means the front of the tire swings away from the car. Read the angle off the turn plate scale, not by eye.
- Turn the wheel 20 degrees in and read the KPI vial. The bubble has moved through a full 40 degree sweep, and the number it lands on is your kingpin inclination.
Then do the other side, and compare. Left and right should agree within half a degree. A gap above one degree between the two sides is a finding in itself, and section five covers what to do with it.
The 20 degree figure is the whole method. The gauge is calibrated for that sweep, and a 15 degree swing on one side and 25 on the other gives you two numbers that mean nothing. Turn plates with a readable dial are in the kit for that reason. Chalk marks on the floor will not get you to a true 20.
Four reasons the reading looks wrong
A support call we took in June shows all four at once. A 1976 Dodge 1.5 ton truck read 15 degrees of KPI against a factory figure of 7.6, and the truck drove fine at highway speed. Nobody had bent that axle. Every item on this list was true of that measurement. Work through them in this order, because this is the order they usually turn up in.
1. Camber was not at spec when you measured
This is the most common one by a wide margin, and it is the one people skip because it feels like a separate job. It is not. Whatever camber error is present when you start the sweep comes through in the KPI reading. Set camber, or at least measure it so you know how far off the base is, before you touch the KPI vial.
2. The hub face is corroded, painted, or uneven
The gauge sits on a magnet, and the magnet sits on the hub face or on the clamp's adapter plate. Flaking rust, a thick coat of paint, or a burr from a lug stud tilts the whole gauge a degree or two before you have measured anything. On a vehicle that has been outside for a few decades this is nearly guaranteed.
Take the wheel off, run a wire brush over the hub face until bare metal shows, and wipe it down. If the face is pitted past cleaning, mount the gauge on the wheel clamp instead. The clamp holds its adapter plate flat to the rim, which gives the magnet a true surface no matter what the hub looks like.
3. The kingpin is worn
Solid axle trucks and older cars steer on a kingpin running through bushings. When those bushings wear, the knuckle rocks on the pin, and the sweep from 20 out to 20 in moves the knuckle through that slop as well as through its steering arc. The gauge adds the slop to the reading.
Check for it before you measure. With the axle supported and the tire off the ground, grab the tire at the top and bottom and rock it. Movement you can feel at the top and bottom, with the wheel bearing tight, is kingpin play. A worn kingpin will never give a repeatable KPI number, and it is a repair in its own right.
4. The sweep did not reach a true 20 degrees
Without plates, most people under-turn. A sweep that stops at 15 degrees each way reads lower than the real angle, an uneven sweep reads whichever way the error fell, and neither one matches the other side. The fix is the plate dial, read at both ends of the sweep, every time.
The Dodge makes the same point as all four items. A reading that is far from spec on a vehicle that drives straight is a measurement problem until proven otherwise.
The included angle, and what it tells you
One piece of arithmetic separates a bent part from a bad measurement.
Included angle = KPI + camber
Both angles are measured from the same vertical line, so adding them gives the angle between the steering axis and the wheel's own centerline. That angle is built into the knuckle. Camber adjusters move the knuckle as a unit, so they change camber and KPI by equal and opposite amounts and leave the included angle where it was. Only a bent or wrong part changes it.
Take the Mustang figure as an example. Factory KPI is 6.87 degrees. Say your camber gauge reads half a degree positive on that side. The included angle is 6.87 plus 0.5, or about 7.4 degrees. Now run the same sum on the other side of the car. The two included angles should land within half a degree of each other.
Read the two results together.
- KPI is off, but the included angle matches the other side. The knuckle is straight. Look at camber, then at the measurement.
- KPI and the included angle are both off on the same side, and the other side is normal. Something between the ball joints is bent, or the wrong knuckle is fitted. No adjustment will bring it back.
That is a diagnosis most people pay a shop for, and it takes a gauge, a calculator, and the plates.
When KPI really is out of spec
Once the four measurement errors are ruled out and the included angle still says the knuckle is bent, the cause is one of three things. A bent steering knuckle or spindle, usually from a curb or a pothole. A worn kingpin on a solid axle. Or a replacement knuckle, spindle, or axle that was never the right part for the vehicle.
Work it in this order.
- Rule out the four errors above, and measure again.
- Compare left to right. Under half a degree apart is normal. Over one degree is a structural finding on the side that reads wrong.
- Run the included angle on both sides, and note which side is out.
- Take those numbers to a frame or chassis specialist. You are walking in with the answer, not the question.
The gauge did its job either way. On the Dodge, it sent the owner to clean a hub face and check a kingpin instead of buying an axle. On a car that really has a bent knuckle, it names the side and the part before anyone puts it on a lift.
KPI measurement checklist
Keep this open on your phone in the garage.
- Car level, tires at pressure, steering wheel free to turn
- Camber measured and set to factory spec on both sides
- Hub face wire-brushed to bare metal, or gauge on the wheel clamp's adapter plate
- Kingpin or ball joint play checked with the wheel rocked top to bottom
- Turn plates under both front wheels, dials at zero with the wheels straight
- Wheel turned 20 degrees out on the dial, KPI vial zeroed
- Wheel turned 20 degrees in on the dial, KPI read
- Other side measured the same way
- Left and right within half a degree of each other
- Included angle worked out as KPI plus camber on each side, and compared
A reading that still looks wrong after all ten lines is telling you something true about the car. Everything before that is telling you something about the setup.