Collars vs Harnesses. Why the welfare conversation misses the real issue
- Eryn Martyn-Godfrey
- Dec 11, 2025
- 4 min read
Updated: Dec 19, 2025
There has been a growing trend in the dog training world to publicly bad-mouth collars while positioning harnesses as the ethical or welfare-friendly alternative. I want to pause that narrative, because it oversimplifies a complex issue and risks replacing one welfare myth with another.
Yes, the canine neck contains vital and delicate structures, and excessive pressure, repeated jerks, or sustained tension on the lead are not acceptable. That is not in dispute. What is far less often discussed is that dogs are quadrupeds whose movement and propulsion come primarily from the forequarters and shoulders. When a dog pulls, force is not removed by a harness. It is redistributed.
Most commonly used harness designs load force across the scapula, sternum, ribs and thoracic spine. The canine shoulder is not a fixed joint but a muscular sling that relies on free movement for normal stride length. Research and gait analysis have shown that even well-fitted harnesses can restrict shoulder extension and alter stride length, particularly when a dog is pulling. Changes in gait are not neutral and may increase compensatory loading elsewhere in the body.
Studies frequently cited to condemn collars often measure peak pressure under artificial conditions using rigid models rather than living dogs. They do not compare this with harness-related pressure, harness-induced gait changes, or long-term musculoskeletal impact. Even within this research, the authors consistently emphasise that the core issue is pulling itself and that dogs should be trained to walk on a loose lead, with equipment discussed as management rather than a welfare solution.
What concerns me most is that equipment-focused messaging distracts from the actual causes of the behaviour. Dogs do not pull because they are wearing a collar. They pull because of arousal, reinforcement history, frustration, fear, pain, or unmet needs. A harness does not teach loose lead walking and can increase mechanical advantage, allowing dogs to pull more effectively using the chest and shoulders. A flat collar used with skill, appropriate reinforcement, and a dog trained to maintain slack in the lead may involve minimal pressure on the neck at all.
As trainers and behaviour professionals, we need to be careful about the narratives we promote. Demonising collars while presenting harnesses as inherently safe encourages fear-based decisions rather than informed, individualised choice. Welfare is not about choosing the right piece of equipment. It is about reducing tension, force and conflict through training, emotional regulation, and addressing why the behaviour exists in the first place.
A truly welfare-led approach is not collar versus harness. It is training first, assessment first, and skill building first, with equipment selected thoughtfully for the individual dog and used as management, not as a substitute for teaching.
If you are dealing with lead pulling and you want help that is practical, kind, and evidence-led, you can work with me in three ways.
Book a 1:1 training session for a tailored loose lead plan.
If you suspect pain or discomfort is part of the picture, I can help you gather clear observations and build a referral-ready summary for your vet. Click here
If you are a trainer and want to deepen your skills in observation, biomechanics, and welfare-led decision making, explore my mentoring and CPD options. https://www.believeinmagic.dog/dog-training-professional-mentoring
References and further reading
1. Pauli, Bentley, Diehl & Miller (2006) – Collar vs Harness intraocular pressure
It’s not freely open-access but is the authoritative index for the article.
Pauli AM, Bentley E, Diehl KA, Miller PE. Effects of the application of neck pressure by a collar or harness on intraocular pressure in dogs. J Am Anim Hosp Assoc. 2006;42(3):207–211.
2. Carter, McNally & Roshier (2020) – Collar pressure on a model neck
This is the PubMed entry for the Nottingham Trent University study on collar pressures:
Anne Carter, Donal McNally, Amanda Roshier. Canine collars: an investigation of collar type and the forces applied to a simulated neck model. Vet Rec. 2020;187(7):e52.
If you need the institutional PDF, this repository version is available:
3. Lafuente, Provis & Schmalz (2018/2019) – Harness effects on shoulder extension
This PubMed entry gives citation details (including DOI) — full text usually requires journal access:
M Pilar Lafuente, Laura Provis, Emily Anne Schmalz. Effects of restrictive and non-restrictive harnesses on shoulder extension in dogs at walk and trot. Vet Rec. 2019;184(2):64.
4. Peham C, Limbeck S, Galla K, Bockstahler B (2013) – Harness influence on gait/pressure
This Veterinary Journal article is the likely correct match for the “Peham et al.” study you referenced. It’s not freely open access but does exist and is indexed on Scopus/ScienceDirect:
Peham C, Limbeck S, Galla K, Bockstahler B. Influence of harnesses and head-collars on the kinematics of the canine gait. Vet J. 2013;198(3):e93–e98.
If you don’t have access via ScienceDirect, you may be able to request a PDF via ResearchGate or an institutional library.
5. AVSAB Position Statements (humane training & equipment)
The general resource page for position statements (including humane training and equipment use) from the American Veterinary Society of Animal Behavior (AVSAB):
6. Nottingham Trent University press article (“Collars risk causing neck injuries in dogs”)
If you want the actual research report rather than the press release, the NTU repository (IRep) has the PDF that is accessible:
If you want evidence on pulling / gait mechanics more broadly, there are a few other indexed studies (e.g., Cavalieri review on gait effects with harness and head-collars) that are accessible via PubMed Central.
This post has had a lot of interest on Facebook