A deadlift that slows near lockout is not automatically a hitch, and a bar that brushes the thighs is not automatically a failed lift. Replay the final third of the repetition, identify whether the ascent stays continuous, and separate a training fault from a federation-specific judging question. Then use a small, repeatable regression instead of trying to force a dramatic “hips through” finish.
This guide owns visual diagnosis of hitching, ramping, and lockout changes plus conservative training experiments. It does not diagnose a muscle, joint, nerve, or spine condition; prescribe a maximum-effort attempt; or decide a competition result for every federation. The conventional deadlift exercise guide owns the complete setup and execution tutorial. For a different loading option, see the trap-bar deadlift and deadlift alternatives guides. If a lift causes sharp pain, numbness, weakness, or loss of control, stop and seek an appropriately qualified in-person assessment.
Direct answer: define the event before you fix it
Use these words as observations, not diagnoses:
| Term | Working description for replay | Do not infer |
|---|---|---|
| Slow lockout | Upward movement continues, even if it is very slow, through the final range. | That the lift is automatically legal or technically ideal in a meet. |
| Hitch | The ascent becomes discontinuous: the lifter appears to re-bend the knees, stop the hips, jerk the bar, or use successive upward steps to finish. | A particular weak muscle or an injury from a video. |
| Ramp | The bar travels up the thighs while the torso and knees change position, often giving the appearance of the lifter climbing the bar. | That every federation calls this a failed lift. |
| Ordinary thigh contact | The bar brushes or tracks along the thighs without a visible stop, reverse, or support. | That contact alone proves a fault. |
| Support | The bar’s upward travel is assisted by resting on or being held up by the thighs. | That a phone angle can settle a judging call when the contact is obscured. |
Hitching and ramping may overlap, but they are not synonyms. A ramp can be smooth and continuous yet still raise a rule question if the bar is supported. A hitch can occur without obvious thigh contact if the knees and hips repeatedly change sequence. The useful first question is therefore: what moved, when did it move, and did the bar remain continuously controlled?
Training fault versus meet ruling
Gym coaching and competition judging answer different questions. In training, you can label any loss of continuous lockout as a fault worth reducing because it changes the skill you intended to practise. In a meet, only the federation’s current technical rules and the referees’ view determine whether the attempt receives a decision. A rep can be a useful training repetition and still be ruled no-lift; it can also look awkward in a gym video without matching a rule’s exact disqualification language.
IPF rule boundary — Technical Rulebook effective 01 March 2026, version 3: the deadlift disqualification list includes supporting the bar on the thighs. The same rule text distinguishes a bar that edges up the thighs without being supported, stating that this alone is not a reason for disqualification. It also lists downward movement before the final position, while noting that a bar settling as the shoulders come back is not, by itself, a disqualification. These are IPF claims tied to that dated rulebook, not universal powerlifting rules.1
If you lift under another federation, do not copy that paragraph as your ruling. Open its current rulebook, note the effective date and deadlift section, and ask the meet director or head referee how ambiguous contact will be interpreted. Rulebooks can change; a search snippet, gym folklore, or a clip from another federation is not a substitute for the document governing your platform.
Replay the final third before changing your technique
Film from a true side view, with the whole bar, plates, feet, knees, hips, trunk, and lockout position visible. A rear-quarter view can add context, but it can make forward-and-back movement look different from a side view. Keep shoes, plates, camera position, and load constant while comparing clips. Do not diagnose from the rep that is most dramatic or from one paused frame.
Task asset 1 — the six-frame replay checklist
Pause each repetition at the same landmarks. Use yes, no, or unclear rather than a confident label when the camera cannot answer.
| Frame | Mark the bar and body landmarks | Question to answer |
|---|---|---|
| 1. Knee-height approach | Bar, knees, hips, shoulders | Is the bar still travelling upward without a stop? |
| 2. First thigh contact | Bar-to-thigh relationship | Is contact a brush, a track, or a visible support point? |
| 3. First slowdown | Bar height and torso angle | Does speed drop without the knees changing direction? |
| 4. Lockout attempt | Knees, hips, shoulders, bar | Do the knees and hips continue to extend, or does one reverse? |
| 5. Any second effort | Bar, knees, hips | Is there a jerk, knee re-bend, or repeated upward step? |
| 6. Finish | Bar, shoulders, knees | Is the final position motionless and controlled before the down signal? |
Write the first observable event in one sentence: “Bar slowed but kept rising,” “knees re-bent after thigh contact,” or “contact was obscured.” This sentence is more useful than “my lockout is weak.”
What the bar and knees can tell you
A side video can describe timing and visible relationships: bar height, direction, knee angle changes, hip travel, torso change, and whether a pause or reversal appears. It cannot identify which tissue produced the motion, measure force accurately, or determine the legal outcome under every ruleset. Keep those boundaries in the notes and claim ledger.
Knee and hip sequencing: find the first change
The final third of a deadlift is a sequence, not a single “squeeze.” As the bar passes the knees, look for a broadly continuous extension of the knees and hips while the bar remains controlled. People use different proportions, stances, and lockout strategies, so the target is not an identical angle. The diagnostic question is whether the sequence changes abruptly under the tested condition.
Task asset 2 — the knee/hip sequence card
Review three repetitions and mark the first event that differs from your controlled baseline.
| Event | Rep 1 | Rep 2 | Rep 3 | Interpretation to test |
|---|---|---|---|---|
| Knees continue extending after the bar passes them | If no, inspect for a knee re-bend or bar stop. | |||
| Hips continue toward the standing finish | If no, inspect for a hip stop or torso climb. | |||
| Bar remains on an upward path | If no, mark a pause or downward movement. | |||
| Torso changes without a bar stop | A position change is not automatically a hitch; compare control. | |||
| Torso changes with a bar stop or second effort | A hitch hypothesis becomes stronger. | |||
| Thigh contact is visible but not supporting | Contact alone does not answer a rule question. | |||
| Fatigue changes the sequence on the final rep | Test fewer reps or a lower load before changing technique. |
Do not turn this into a muscle diagnosis. If the first visible event is a knee re-bend, test a manageable load and a controlled knee-to-hip transition. If the bar pauses while the knees and hips remain stable, test load and fatigue first. If the video cannot separate the events, improve the camera angle or ask a qualified coach to review it.
Common patterns and the smallest useful test
A smooth but slow finish
A slow rep can still be continuous. If the bar keeps moving upward, the knees do not re-bend, the hips do not stop and restart, and no support point appears, label it “slow/continuous” rather than “hitch.” For training, decide whether this speed and effort fit the session; then use a load that lets you practise the intended lockout without repeated grinders. There is no need to prescribe or chase a specific percentage here.
Research on sticking regions describes portions of a heavy lift where bar velocity falls and mechanical advantage changes, but the location and expression vary by lift and lifter.2 That supports marking the slow region; it does not convert a slow region into a universal failed-lift rule or establish a personal diagnosis.
Knees re-bend while the bar climbs
If the knees visibly move forward or re-bend after the bar has passed them, mark whether the bar also pauses or whether the torso climbs the bar. Start with a lighter, repeatable test at the same setup. Cue a continuous finish only if that cue changes the sequence without jerking the bar or forcing a new position. If the pattern appears only after several repetitions, shorten the set and retest before changing your entire deadlift.
Hips stop, then the torso leans back
A torso change can be part of a controlled finish; a stop-and-lean or successive upward “step” is the training fault to test. Lower the condition that exposes it, then practise the last part of the pull from a manageable height if you can do so without turning the exercise into a different task. A partial-range drill can rehearse the lockout, but it does not prove the floor-start deadlift is fixed. Return to the full pull and compare the same replay landmarks.
The bar slides up the thighs
First separate contact from support. A bar can stay close and brush the thighs as it travels. A ramping appearance becomes more concerning for training when the bar stops, the lifter climbs around it, or the thighs visibly prop it upward. For a meet, do not self-judge from a side clip if the contact point is obscured; use the federation-specific rulebook and officials’ guidance. Under the dated IPF rule cited above, supported contact and non-supported edging are treated differently.1
Downward movement at the top
Replay the bar and shoulders together. A true downward movement of the bar before the final position is a different observation from a bar that appears to settle as the shoulders come back. The IPF Technical Rulebook effective 01 March 2026 addresses that distinction; other federations may use different language or interpretation.1 In training, reduce the load or stop the set before repeated reversals. Do not practise a deliberate dip just because someone says “ride it down.”
Task asset 3 — the fix ladder
Move down the ladder only as far as needed to regain a repeatable, continuous finish. Keep the camera angle and most variables constant.
| Level | Test | Keep it when | Do not claim |
|---|---|---|---|
| 1. Clean replay | Record three controlled reps and name the first change | You can distinguish slow/continuous from stop/restart | A video label identifies anatomy or a rule result |
| 2. Reduce exposure | Use a lighter or less-fatiguing set at the same setup | The lockout stays continuous and the bar remains controlled | The lighter load proves full-load readiness |
| 3. One cue | Use one cue for knee/hip continuity or bar closeness | The cue improves repeatability without a jerk or new compensation | The cue is universal or medically corrective |
| 4. Lockout drill | Practise a controlled partial-range or pause variation | The drill is repeatable and transfers to the full pull | A partial range is the same as a competition deadlift |
| 5. Full-pull rebuild | Reintroduce the floor-start pull with conservative conditions | The same replay landmarks remain consistent across sessions | You need a max attempt to test the change |
“Conservative” here means a condition you can repeat with control, not a prescribed load or percentage. Log the exact variation, range, repetitions, and observed first change. If a change improves the video only because the lift became a different exercise, record it as a regression rather than a solved lockout.
Task asset 4 — meet or training decision card
Use this card before changing a platform attempt or interpreting a gym rep.
| Question | Training decision | Meet decision |
|---|---|---|
| Is the ascent slow but continuous? | Keep the rep as “slow/continuous,” then adjust the training condition if the skill is not repeatable. | Read the federation’s lockout and downward-movement wording; officials decide the attempt. |
| Is there a stop, jerk, knee re-bend, or second effort? | Treat it as a lockout fault to regress and retest. | Check the current rulebook’s deadlift disqualification section; do not import another federation’s ruling. |
| Does the bar contact the thighs? | Check whether contact changes control or becomes support. | Identify your federation, rulebook version/date, and the exact support/contact language. |
| Is the contact obscured on video? | Re-film from a clearer angle or ask a coach. | Do not claim “good” or “no lift” from a hidden contact point. |
| Does the issue appear only near fatigue? | Reduce repetitions or load and rebuild continuity. | Select attempts you can execute under the meet’s rules; do not use a training grind to discover the boundary. |
| Is there pain, weakness, or loss of control? | Stop and seek appropriate assessment. | Tell the meet staff if safety is affected; do not continue an unsafe attempt. |
The meet column is intentionally narrow. As of the cited IPF Technical Rulebook effective 01 March 2026, support on the thighs and downward movement are explicit deadlift disqualification topics, while non-supported edging is treated separately.1 That sentence should not be copied into an unverified claim about another federation.
A two-session retest log
Use the same camera, shoes, plates, and variation where possible. Change one input per row.
| Session | Variation / range | Load and reps | One variable changed | First event | Continuous? | Thigh contact/support clear? | Next action |
|---|---|---|---|---|---|---|---|
| Baseline | none | ||||||
| Test A | |||||||
| Test B | |||||||
| Transfer | full pull |
Use plain descriptions: “bar slowed, no stop,” “knees re-bent after contact,” “support unclear,” or “continuous at reduced condition.” If the pattern changes only at the heaviest or most fatigued condition, the result is a boundary to programme around—not proof of a structural problem and not a reason to test a maximum.
What not to do
Do not force “hips through” into a lean-back
A cue can be useful when it improves a continuous finish, but an aggressive lean-back can create a new torso change that looks like the problem you were trying to remove. Use one cue, inspect the bar and knee/hip sequence, and reject it if control worsens.
Do not call every thigh brush a hitch
The bar’s proximity to the legs is a useful observation. It becomes a training question when it interrupts the ascent, changes the knee/hip sequence, or requires support. It becomes a competition question only in the context of a named federation’s current rules.
Do not use a lockout drill as a diagnosis
Block pulls, pauses, and tempo work can expose or reduce a timing problem. They cannot identify an injured tissue or prove why a floor pull changed. Use them as task-specific training experiments and return to the full movement for transfer.
Do not let a referee rule become a coaching rule
A judged no-lift is an event under one rule set. It is not automatic evidence of injury, weakness, or a universal technique error. Conversely, a gym rep that receives no comment is not proof that it would pass in a particular federation.
When to get another set of eyes
Ask a qualified coach to review the clip when the camera cannot show the bar-to-thigh relationship, the first change differs across repetitions, or you cannot tell a pause from a perspective artifact. Bring the original clip, load, variation, repetition number, and the one-variable retest log.
Stop the set for sharp or escalating pain, new numbness or tingling, weakness, loss of balance, or loss of control. Seek appropriately qualified clinical advice for persistent or worsening symptoms. This guide can organise observations and training choices; it cannot diagnose or treat a condition.
Direct answer
Replay the final third of the deadlift and classify the event: slow but continuous, hitching, ramping, ordinary thigh contact, or support/unclear. Then find the first knee, hip, torso, or bar change and use the smallest rung of the fix ladder that restores continuous control. For competition, stop at the boundary: cite the exact federation and dated rulebook, because IPF’s 01 March 2026 rules are not a universal standard. A clean training regression is valuable even when it cannot predict a referee’s decision.
Claim ledger
| Claim | Evidence type | Confidence and boundary |
|---|---|---|
| Hitching, ramping, ordinary thigh contact, and support are distinct replay descriptions that may overlap | Defined coaching vocabulary and movement-observation framework | High as an operational distinction; labels do not diagnose anatomy or settle a federation ruling |
| A slow but continuous lockout is different from a stop, reversal, or repeated upward effort | Movement replay method | High as a visible distinction when camera angle and frame rate are adequate; video cannot measure force reliably |
| Sticking regions are phases where bar velocity can fall during heavy lifts | Review of sticking-point research, Kompf and Arandjelović2 | Moderate; location and expression vary by lifter, lift, and protocol |
| Under the IPF Technical Rulebook effective 01 March 2026, supporting the bar on the thighs is listed as a deadlift disqualification, while non-supported edging is treated differently | Federation rulebook, dated 01 March 2026, version 31 | High for the cited IPF ruleset; not transferable to another federation or future edition |
| Under that same IPF rulebook, downward bar movement before the final position is listed, with a separate note about settling as the shoulders come back | Federation rulebook, dated 01 March 2026, version 31 | High for the cited IPF ruleset; officials and the full rule text govern a meet |
| Reducing load, fatigue, range, or complexity and changing one variable at a time is a useful way to test a training hypothesis | Conservative coaching method | Practical guidance, not a validated diagnosis or a prescription for a particular load |
| A replay cannot identify a muscle, joint, nerve, or spine condition | Scope boundary | High; individual diagnosis requires appropriate assessment |
Footnotes
-
International Powerlifting Federation. IPF Technical Rulebook, effective 01 March 2026, version 3, Deadlift rules and causes for disqualification. Accessed 2026-08-29. The cited statements are specific to this dated IPF document. ↩ ↩2 ↩3 ↩4 ↩5 ↩6
-
Kompf J, Arandjelović O. The sticking point in the bench press, the squat, and the deadlift. Sports Medicine. 2016;47:631–640. ↩ ↩2
Frequently Asked Questions
- What is hitching in a deadlift?
- Hitching is a coaching description for an interrupted ascent in which the bar or lifter loses a continuous lockout strategy, often with a knee re-bend, hip stop, torso change, or repeated upward jerks. A slow but continuous pull is not automatically a hitch. Use replay to identify the first change rather than diagnosing from a single frame.
- Is ramping the same as hitching?
- No. Ramping is a description of the bar travelling or sliding upward along the thighs while the lifter changes position; hitching describes an interrupted or discontinuous ascent. They can appear together, but neither word is a universal federation ruling. Check your own federation's current rulebook for how support, downward movement, and lockout are judged.
- Is it okay for the bar to touch my thighs during a deadlift?
- Light contact can occur as the bar passes the legs and is not, by itself, proof of a failed lift or a training fault. The important observations are whether the bar is supported, whether its upward travel stops or reverses, and whether the knees and hips keep sequencing. Competition rules differ, so do not turn ordinary gym contact into a universal meet ruling.
- How do I fix a deadlift hitch at lockout?
- First film the final third of the lift and lower the test load or shorten the set so you can identify the first loss of continuous motion. Then use the smallest regression that matches the observation: practise a controlled lockout from a manageable height, use paused or tempo work, or rebuild the full pull with a load that does not require a hitch. Change one variable at a time.
- Can I hitch a deadlift in a powerlifting meet?
- There is no universal answer because federations publish different rules and referees apply the rulebook for that meet. Under the IPF Technical Rulebook effective 01 March 2026, supporting the bar on the thighs is a listed deadlift disqualification, while a bar that edges up the thighs without being supported is expressly treated differently; downward movement before the final position is also listed. Read your federation's dated rulebook and briefing.
- Should I train through a hitch if the lift is not painful?
- A painless hitch is still a useful training signal, not a reason to force repeated failed-looking reps. Reduce the condition that exposes it, keep a repeatable lockout, and log whether the change appears only under fatigue or load. Stop for sharp or escalating pain, numbness, weakness, loss of balance, or loss of control and seek qualified assessment.