Bottom Bracket Creaking Under Load — How to Find and Fix It

Bottom bracket creaking under load is one of the most maddening sounds in cycling. It mocks you on every climb, every hard sprint, every moment you’re putting down serious watts. I’ve spent an embarrassing number of Saturday mornings chasing this exact noise across three different bikes — and the thing that consistently wasted my time was skipping straight to the fix before confirming the source. This article is built around a different approach: systematic isolation first, then targeted repair based on your specific BB standard. If you read nothing else, read the first section before you touch a wrench.

Is It Actually the Bottom Bracket — The 3-Minute Test

Most creak diagnoses fail at step zero. People assume the noise is the BB because it happens when they pedal. The problem is that pedaling loads about six different contact points simultaneously — the pedals, the cranks, the BB shell, the chainring bolts, the seatpost, and even the saddle rails. All of them can produce a rhythmic creak that sounds like it’s coming from directly below your groin.

Here’s the quick isolation method I use before doing anything else.

The Seated vs Standing Test

Ride normally and note when the creak occurs. Then do two specific tests back to back:

Seated, low effort: Spin at 90 RPM on flat ground with minimal pressure. No creak? Good.
Standing, hard effort: Get out of the saddle and stomp on the pedals on a moderate climb. Creak returns?

If the creak only appears under load — especially standing — and disappears when spinning lightly seated, you’ve narrowed it down significantly. That pattern points to the BB or cranks. A seatpost creak, by contrast, usually appears seated and under load but often disappears entirely when you stand. Saddle rail creaks are almost exclusively seated.

The One-Leg Test

This takes about 90 seconds. Pedal using only your right leg, keeping the left foot off the pedal entirely. Listen. Then switch. If the creak only appears on one side, you’re likely looking at a pedal thread issue or a crank arm bolt on that side — not the BB itself. A true BB creak typically fires on both sides because both cranks load the same spindle.

The Stationary Stress Test

Lean the bike against a wall. Grab each crank arm individually and push and pull axially — meaning along the direction the spindle runs, not rotationally. Any click or movement means your preload is off. Then grab both crank arms and try to wiggle them laterally while loading the pedals by hand. A creak or click here is almost certainly the BB interface or the crank/spindle junction.

Probably should have opened with this section, honestly. Half the BB overhauls I’ve done over the years were unnecessary — the noise was a pedal spindle or a loose chainring bolt, and ten minutes with a 5mm hex key would have solved it.

Threaded BB Creak — Remove, Clean, Retorque

BSA/English threaded bottom brackets — the standard with a 68mm or 73mm shell, right-hand drive side threading, left-hand non-drive side threading — are the most forgiving to work with. The fix is almost always the same: remove, clean, apply fresh lubrication, reinstall to correct torque.

Removal and Inspection

You’ll need a BB tool specific to your cup style. For a Shimano Hollowtech II external cup BB like the SM-BB52 or SM-BB72, that’s the TL-FC32 or equivalent 16-notch tool. For a square taper ISIS cup, you’ll want a splined socket. Drop your chain off the chainring before starting.

Remove both cups. When you pull them out, look at the threads in the shell. If you see dry, bare metal threads or aluminum oxide deposits (a grey-white powder, common on alloy frames with steel cups), that’s your creak. The micro-movement between the cup threads and the shell under pedaling load generates that sound — and it’s incredibly effective at projecting through the frame to sound like it’s everywhere at once.

Grease vs Anti-Seize — The Clear Answer

Use grease on aluminum or carbon frames. Use anti-seize on steel frames. Here’s the reasoning: on aluminum frames, the bigger threat is galvanic corrosion between the steel or titanium cup and the alloy shell — grease handles this well and provides enough lubrication to prevent creak. On steel frames, corrosion is a slower process but thread galling is a real risk if you’re reinstalling steel cups into a steel shell repeatedly. Anti-seize — something like Loctite 771-64 or Finish Line Anti-Seize — reduces that risk and still eliminates creak.

Don’t use threadlocker (like blue Loctite 243) on BB threads. I made that mistake once on a Trek Émonda and spent 45 minutes heating the shell with a heat gun to get the cups out six months later. Not worth it.

Torque Specs

Reinstall both cups by hand until snug, then torque to spec:

Drive-side cup (right side, reverse threaded): 35–50 Nm
Non-drive cup (left side, standard threaded): 35–50 Nm
Shimano external cups specifically: 35–50 Nm per their technical documentation

If your frame is carbon, stay at the lower end — 35 Nm. If you’re on steel or aluminum and the creak has been persistent, going to 50 Nm is fine. Go back and retorque after the first hard ride. Threaded BBs settle slightly under initial load.

Press-Fit BB Creak — The Notorious Fix

Press-fit bottom brackets are the reason entire Facebook groups exist. BB86, BB90, BB92, PF30, BB386 — they all share the same fundamental vulnerability: a plastic or aluminum cup pressed into a composite or aluminum shell with no mechanical fastener. Under load, the cup can micro-move against the shell. That’s the creak.

Frustrated by a 2019 Trek Domane with a BB86 that creaked despite three separate shop visits, I eventually learned the actual fix sequence rather than just spraying carbon paste and hoping. Here’s what works.

The Loctite 609 Method

Loctite 609 is a retaining compound — not threadlocker, not adhesive in the traditional sense. It fills micro-gaps between pressed components and cures to eliminate the movement that causes creak. It is the go-to fix for press-fit BB creak when the cups themselves are otherwise in good condition.

The process:

Remove the crankset and press out both BB cups using a press tool or a dead-blow hammer with a proper drift — never improvise this with a screwdriver or you’ll damage the shell.
Clean the shell bore thoroughly with isopropyl alcohol. Clean the outer cup surface the same way. Let both dry completely — 609 needs clean, dry surfaces to cure properly.
Apply a thin, even coat of Loctite 609 to the outer diameter of each cup. Not dripping wet — just a uniform film.
Press the cups back in using a proper BB press tool. The Wheels Manufacturing BB Press Tool is about $120 and worth every cent if you’re doing this yourself more than once.
Let cure for 24 hours before riding.

This fixes the creak in roughly 80% of press-fit cases in my experience.

When Shims Are the Answer

If the shell bore is oversized from repeated press-fit installations — which happens, especially on older carbon frames — the cups won’t seat tightly even with 609. Wheels Manufacturing makes BB shell shims for exactly this situation. The WM-PF30-SHIM set adds 0.5mm to the bore depth and restores the interference fit. Measure your bore diameter with a set of calipers before ordering — standard PF30 bore is 46mm, BB86 is 41mm.

When to Convert to Threaded

Sometimes the right move is to abandon press-fit entirely. If your frame is BB86 or BB92, you can convert to a threaded T47 interface using an insert system like the Wheels Manufacturing PF86/92 to T47 conversion. This is a permanent fix that eliminates all future creak potential for that frame. It costs around $80–120 for the conversion kit plus whatever T47 BB you run. For frames that have seen four or more press-fit BB replacements, this is the most cost-effective long-term solution.

PF30 frames can convert to 30mm spindle threaded BBs using similar kits, or — with the right adapter — even run a 24mm spindle if you want to run Shimano cranks. That’s a whole separate decision tree, but the creak-elimination benefit is the same.

T47 BB Creak — The Rare Case

T47 is the large-diameter threaded standard that was supposed to solve everything, and largely it has. Creak is rare with T47. When it does occur, the fix is almost always one of three things.

Check Preload First

T47 external cup designs like the Chris King T47 or Wheels Manufacturing T47 external use a preload system similar to traditional threaded BBs. If the preload cap is loose or if the crank arm bolt wasn’t torqued correctly after installation, you’ll get movement at the crank/spindle interface that sounds like a BB creak but isn’t. Check crank arm bolt torque — Shimano cranks use 12–14 Nm, SRAM DUB cranks use 40–54 Nm depending on the model. Those numbers are very different. Using the wrong spec is a common mistake.

Clean and Re-Grease the Threads

T47 uses standard right-hand threading on both sides (the major improvement over BSA). Removal and reinstallation follow the same logic as threaded BBs above — clean shell threads, apply grease, retorque to 40–50 Nm. Because T47 shells are typically larger diameter and better supported, creak is much less common, but the fix when it happens is identical to the BSA process.

Creaks That Are Not the BB at All

Here’s where I save you a lot of time. A significant percentage of “BB creaks” are not the BB. Run through these before committing to a BB overhaul.

Pedal Threads

Remove both pedals. Clean the pedal threads with a rag. Apply a proper grease — not anti-seize here, grease — and reinstall to 35 Nm. Dry pedal threads creak under load and the sound projects directly into the crank arm. This is the single most common false-positive in creak diagnosis.

Chainring Bolts

Use a 5mm hex key and snug all chainring bolts. On single chainring setups with a direct mount spider, torque to 8–10 Nm. On traditional 5-bolt rings, 6–8 Nm. Loose chainring bolts produce a rhythmic creak that is nearly impossible to distinguish from a BB creak by sound alone.

Seatpost Clamp

Pull the seatpost out completely. Clean the post, apply carbon paste if it’s a carbon post or carbon frame, and reinstall. Torque the clamp to spec — typically 5–6 Nm for a single-bolt clamp, 4 Nm each for dual-bolt designs. A seatpost creak usually appears seated and under load, which overlaps with the BB creak profile enough to fool most people.

Saddle Rails

The rails where they contact the saddle clamp can creak under rider weight. Remove the saddle, clean the rail contact surfaces, apply a thin layer of grease, and reinstall. If you have a carbon-railed saddle like the Fizik Antares R1, this is a surprisingly common culprit.

Headset

Less common as a “pedaling creak” but worth ruling out. A loose headset clicks and creaks under braking and cornering more than pedaling, but on climbs where you’re pulling on the bars, it can mimic a BB creak. Check headset preload by squeezing the front brake and rocking the bike fore-aft. Any play, click, or clunk means a headset preload adjustment is needed before you go further.

Work through this list before you pull the cranks and BB. More often than not, the answer is a greased pedal thread and five minutes with a torque wrench — not a full BB replacement. The systematic approach isn’t the exciting answer, but it’s the right one.