What does correlation mean in this context?

It means the module is comparing two things against each other rather than checking either one against a limit. The cruise switches report on two separate circuits, and the module knows which combinations of those two are physically possible. A correlation fault is stored when it receives a combination that cannot happen — two conflicting functions asserted at once, for instance. Neither circuit needs to be out of range for that to occur. The problem is the pair, and that is why this code exists separately from the high, low and range codes on the same switches.

Both wires tested fine. How can there still be a fault?

Because the fault is between them rather than in either one. Two conductors that have been crushed together, or bridged by corrosion in a shared connector, will each remain perfectly continuous from end to end and perfectly isolated from ground — so continuity tests pass, resistance to ground passes, and voltage at the module looks right. What fails is the insulation separating them, and the only measurement that sees it is insulation resistance from one signal wire directly to the other with both ends disconnected. It is a test that has to be chosen deliberately.

This started right after my airbag was replaced. Is that related?

Very probably, and it is the first thing to raise. The two circuits involved run together in one bundle up the column and across the clockspring, and the moment a bundle gets crushed is almost always the moment it is put back together — trapped under a shroud, pinched by a clamp, or forced against an edge on reassembly. This code appears after work behind the steering wheel more often than for any other reason. The part that was fitted is likely fine; the routing of the loom around it is not. Ask whoever did the work to check the harness path before anything else is purchased.

Why does the whole system shut down instead of ignoring one bad input?

Because there is no safe way to choose between two conflicting requests. The two-circuit arrangement exists precisely so that button presses can be verified rather than trusted, since cruise control is the one convenience system that can open the throttle without the driver's foot. When the module receives a combination that cannot be genuine, it has no basis for deciding which half to believe, and acting on the wrong one would mean accelerating or cancelling against the driver's intent. Locking the system out until the contradiction is resolved is the designed response and it is the right one.

OBD-II trouble code

P0585: Cruise Control Multi-Function Input "A"/"B" Correlation

Both circuits can measure perfectly and the fault still be real, because the fault is between them. This is the one code here that a wire-by-wire continuity test cannot find — every conductor is intact end to end.

Low severityPowertrainCruise ControlDrivable short-term

Quick facts

System: Powertrain
Category: Cruise Control
Severity: Low severity
Drivable: Usually safe to drive short-term
Repair cost range: $90 – $550
DIY difficulty: Intermediate DIY

What does P0585 mean?

P0585 is a cross-check failure rather than a circuit failure, and that distinction is the whole page. The cruise control switches report to the module on two separate input circuits, "A" and "B". Each circuit carries its own set of functions, and the module knows which combinations of the two are physically possible. When it receives a combination that cannot occur — accelerate and coast asserted at the same time, or set and cancel simultaneously — it stores this code and shuts cruise control down.

The reason the design works this way is worth stating, because it explains why the module reacts so decisively. Cruise control is the one convenience system on a conventional vehicle that can open the throttle without the driver's foot. Manufacturers therefore build the button reading to be verifiable rather than trusted: two independent circuits whose combined state can be checked for sense, so that a single failure produces a detectable contradiction instead of a plausible-looking false command. When the contradiction appears, locking the system out entirely is the correct behaviour, not an overreaction. There is no safe way to guess which of two conflicting requests the driver meant.

What makes this code genuinely different to diagnose is that both circuits can be in perfect health individually. Circuit A is continuous from end to end and properly isolated from ground. So is circuit B. Every test in the standard repertoire — continuity, resistance to ground, voltage at the module, wiggle testing each wire — will pass. The fault is not in either conductor. It is in the insulation between them. Two wires that have been crushed together, or bridged by corrosion or residue where they arrive side by side in a connector, will carry each other's signal while remaining individually flawless. The measurement that finds it is insulation resistance from one signal wire to the other, with both ends disconnected, and it is not a test anybody performs by accident.

There is a strong pattern in when this code appears, and it is the most useful thing to know before spending money. The two circuits run together in one bundle up through the column and across the clockspring, and the places where a bundle gets crushed are the places where it is clamped, routed through a hole, or trapped during reassembly. That means this code shows up disproportionately soon after somebody has had the steering wheel, the airbag, the clockspring or the column shroud apart. If the fault started within days of that work, the pinch is almost certainly the cause and the part that was fitted is almost certainly fine — which is worth establishing before a second new part is ordered to fix the first one.

A final clarification on the code's title, because it differs from the rest of this block. Elsewhere in the cruise control range the letter identifies a device — multi-function input "A" as opposed to some other switch. Here "A" and "B" are the two halves of one cross-checked pair, and the code is about the relationship between them rather than about either one. It is the only code in this part of the library whose designation refers to a pair rather than to a single thing.

Common causes

Two signal conductors crushed together in the steering column loom
Loom pinched during steering wheel, airbag, clockspring or column reassembly
Corrosion bridging adjacent pins in a shared connector body
Spill residue that has run down the column linking two terminals
Migrated or backed-out terminal touching its neighbour in the connector cavity
Failed switch assembly with an internal short between two button groups
Damaged clockspring conductors shorting to each other rather than to ground
Previous harness repair that spliced two circuits into contact

Symptoms

Cruise control locked out entirely and refusing to engage
Fault appearing within days of steering wheel, airbag or column work
Scan tool showing two cruise functions reported active simultaneously
Multiple cruise buttons appearing to do the same thing
Steering wheel controls for other functions also behaving oddly
Cruise control cancelling immediately every time it is set
Check engine light with normal engine operation in every other respect
Both input circuits testing perfectly good when checked individually

Diagnostic steps

1.Ask when the fault started, and specifically whether anything behind the steering wheel has been apart. This code appears immediately after steering wheel, airbag, clockspring or column work more often than any other in the family, because the fault is a loom pinched on reassembly and the pinch happens at the moment of putting it back together. If a switch pad was recently replaced and this code followed, the new part is almost certainly innocent.
2.Read both input parameters side by side on live data and note which pair of states the module is objecting to. The tool will usually show two independent circuits, and what you want is the impossible combination — two functions reported active at once, or a combination the switch is not physically capable of producing. Knowing which two are involved tells you which two conductors to look at, and that is most of the work.
3.Test insulation resistance between the two signal wires, not from each wire to ground and not end to end. This is the crucial distinction on this code: both conductors are almost always perfectly continuous and perfectly isolated from ground, and every test that checks those two things will pass. The fault lives in the insulation between them, so measure wire to wire with both ends disconnected.
4.Disconnect the switch assembly and repeat the wire-to-wire measurement on the harness alone. If the two circuits are bridged with the switch removed, the fault is in the loom and no switch will fix it. If they only bridge with the switch connected, the assembly is the fault. This is the same fork as elsewhere in the family, but here it is the only way to place the fault, because neither circuit will fail a conventional test on its own.
5.Inspect the loom at every point where it is clamped, routed through a hole or passes a moving part. Look for a flat spot in the outer sheath, a witness mark from a fastener, or a section that has been forced against an edge. Two conductors in one bundle that have been crushed together is the physical form this fault takes, and it will be visible if you get to the right place — which is why the reassembly history matters so much.
6.Check inside the connectors for anything bridging adjacent pins before condemning wiring. Two circuits arriving side by side in one connector body can be linked by corrosion, by residue from a spill that ran down the column, or by a terminal that has migrated in its cavity and is touching its neighbour. Look at the pin faces in good light with the connector split apart rather than probing through it.

Repair cost

$90 – $550

Diagnosis is the larger share here at $100 to $200, because the test that finds this fault — insulation resistance between the two signal conductors — is not part of a routine circuit check and the fault will pass everything else. Repairing a pinched or crushed loom is $80 to $250 of labour with minimal parts, and correcting a bridged connector is often less. A switch assembly with an internal short between button groups is $60 to $350 for the part plus half an hour to two hours. A clockspring is $40 to $250 plus roughly half an hour to two hours including the airbag procedure. If the code followed recent work behind the wheel, raise the reassembly with whoever performed it before buying anything.

Estimate your repair

Run the numbers for your vehicle

Open the Repair Cost Estimator with cruise control switch / stalk replacement preselected. Adjust labor rate and vehicle category to fit your situation.

Estimate your repair cost

DIY vs shop

This is an intermediate DIY job. It usually involves diagnostic steps, specialty parts, and some careful work in tight spaces. If you have the tools and a service manual or trustworthy video for your specific vehicle, it is achievable in a weekend. Otherwise, a competent independent shop will be faster.