The problem
Across independent military and veteran samples, disturbed sleep and trauma-related nightmares are closer to the rule than the exception. The figures below are drawn from peer-reviewed studies, each with its own population and its own measurement threshold — which is why they differ, and why the pattern across them matters more than any single number.
of recent combat veterans with at least subthreshold PTSD presented with insomnia; 61% endorsed distressing nightmares.
Pigeon et al. 2013 · n = 80
of treatment-seeking Canadian Armed Forces members and veterans reported trauma-related nightmares; 68% endorsed sleep disturbances.
Richardson et al. 2018 · n = 663
of US National Guard personnel reported nightmares — 32% nightmares alone, 12% with insomnia as well.
Pruiksma et al. 2021 · n = 841
of their time is how much veterans with a PTSD service dog spent in proximity to that dog, measured by Bluetooth rather than self-report.
Jensen et al. 2022 · n = 82
The clinical significance of those numbers is not simply that veterans sleep badly. It is that insomnia and nightmares behave differently from the rest of the syndrome. In the Pigeon cohort, insomnia was associated with significantly higher PTSD and depression severity at both baseline and six months, and over that interval the prevalence and severity of nightmares diminished only modestly while insomnia severity barely moved at all.¹ The authors drew the obvious inference: addressing sleep early may be a way to alter the course of PTSD rather than merely to make it more bearable.
There is also a specific mechanism worth naming, because it is the one a dog is positioned to act on. Fear of sleep — the anticipatory dread of lying down, of losing vigilance, of what arrives once you do — is measurable, and it accounts for a substantial share of the relationship between a service dog and PTSD severity. That finding is the subject of the next section.
The evidence
This is the most important thing on the page, and it cuts both ways. In the only study to date that measured both sides, veterans partnered with service dogs reported substantially better sleep, far less fear of sleep, and roughly half the odds of nightmares — while objective actigraphy found no significant difference in how long they slept or how efficiently. Any program that tells you otherwise is overselling.
Service dog partnership and sleep in veterans with PTSD, at three months
Adjusted differences between veterans partnered with a service dog (n = 81) and veterans on the provider’s waiting list (n = 74), with 95% confidence intervals. Each measure is on its own scale, so each row has its own axis; the vertical rule marks no difference.
Sleep qualityPSQI global · lower is better
−2.2 (−3.4, −1.1)
Sleep disturbancePROMIS · lower is better
−3.6 (−6.3, −0.9)
Fear of sleepFoSI-SF · lower is better
−6.6 (−9.7, −3.5)
Felt safer falling asleepsingle item · higher is better
+0.7 (0.4, 1.0)
Nightmaresodds ratio · below 1 is better
0.45 (0.26, 0.76)
Nightmare awakeningsper night · lower is better
−0.1 (−0.2, −0.0)
| Measure | Estimate | 95% CI | p | Direction |
|---|---|---|---|---|
| Sleep quality (PSQI global) | −2.2 | −3.4 to −1.1 | <0.001 | Improved |
| Sleep disturbance (PROMIS) | −3.6 | −6.3 to −0.9 | 0.004 | Improved |
| Fear of sleep (FoSI-SF) | −6.6 | −9.7 to −3.5 | <0.001 | Improved |
| Felt safer falling asleep | +0.7 | 0.4 to 1.0 | <0.001 | Improved |
| Nightmares (odds ratio) | 0.45 | 0.26 to 0.76 | 0.003 | Improved |
| Nightmare awakenings per night | −0.1 | −0.2 to −0.0 | 0.007 | Improved |
| Actigraphy: duration, onset latency, wake after sleep onset, efficiency | — | — | n.s. | No significant difference |
Bristol S, Leighton SC, Schwichtenberg AJ, Campbell RL, Ashbeck EL, Taylor DJ, Bedrick EJ, O’Haire ME. The influence of service dog partnerships on perceived and objective sleep quality for military veterans with PTSD. Frontiers in Sleep, 2024. doi:10.3389/frsle.2024.1432919 — a non-randomized trial with waitlist control; allocation followed waiting-list position, participants came from a single service dog provider, and follow-up was three months. Those limitations are the authors’ own.
Why the gap between the two kinds of measurement is the finding
It would be easy to treat the null actigraphy result as a disappointment, or to leave it off a page like this one. It is more useful read the other way. The same study found that fear of sleep accounted for roughly 26% of the service dog’s association with PTSD severity. The dog is not functioning as a sedative. It is functioning as a sentry, and what changes is the handler’s appraisal of the night — whether lying down is safe, whether a nightmare will run to its end unattended, whether anyone will notice.
That is a real mechanism with a real effect size, and it is also a bounded one. A veteran who needs sleep architecture repaired needs the treatments built for that, and the evidence base for those is older and stronger. A randomized trial of cognitive behavioral therapy for insomnia with adjunctive imagery rehearsal therapy in Afghanistan and Iraq veterans with PTSD improved both subjectively and objectively measured sleep and reduced PTSD symptom severity.⁵ The honest position is that these are complements, not substitutes.
The broader mental health evidence
On PTSD symptoms generally, the largest trial to date compared 81 veterans partnered with psychiatric service dogs against 75 on the waiting list, with blinded clinicians administering the assessments. At three months the service dog group scored lower on the self-reported PTSD Checklist for DSM-5 — 41.9 against 51.7, a difference of 11.5 points (95% CI −16.2 to −6.6) — and lower on the clinician-administered CAPS-5, 30.2 against 36.9, a difference of 7.0 points (95% CI −10.8 to −4.5).⁴ Depression, anxiety and several psychosocial measures moved in the same direction.
This trial was also non-randomized, and its authors say so plainly. A randomized trial is underway. Until it reports, the correct description of this intervention is promising and well-measured rather than proven.
Trained tasks
A dog is a service animal under federal law because of its trained tasks, not its temperament, its vest or its paperwork. Two studies asked veterans directly which tasks carried the most weight. The answers were not the ones most people would guess.
Most important · most used
Calming and interrupting anxiety
Across 217 post-9/11 veterans, the trained tasks of calming and interrupting anxiety were rated the most important for their PTSD, used most often in a typical day, and helpful to the largest number of symptoms.⁶
Nighttime
Nightmare interruption
The dog is trained to recognise distress during sleep — thrashing, vocalising, changes in breathing — and to wake the handler by nudging, pawing or licking, then to remain in contact until the handler is oriented and calm.
Hypervigilance
Cover and watch the back
Positioning behind or beside the handler and signalling approach from behind. More frequent use of this task was associated with lower PTSD symptom severity, though with greater anxiety — the kind of mixed result worth knowing about.⁷
Grounding
Deep pressure and physical contact
Applying body weight across the lap or chest to interrupt a panic response or dissociation. In regulatory terms this is work directly related to the disability, which is what makes the dog a service animal.
The finding that should temper everyone’s expectations
In the same survey of 217 veterans, the dogs’ untrained behaviours were rated, on average, as more important to the handler’s PTSD than the trained tasks were.⁶ Veterans still on the waiting list also held higher expectations of task use and importance than veterans who actually had a dog described in practice. A program that promises a list of tasks is describing the smaller half of what it delivers, and a program that lets a waiting veteran believe the dog will be a cure is setting up a disappointment.
One more result belongs here because it bears on how a dog should be raised. Among 82 veteran–dog teams, better mental health at follow-up was associated with less use of dominance-based training methods and closer veteran–dog relationships, and lower dog excitability was associated with less PTSD symptom severity.⁷ Congress reached a compatible conclusion by a different route: the PAWS for Veterans Therapy Act requires that organisations partnering with VA under its pilot program agree not to use shock collars or prong collars and to use positive reinforcement training.¹¹
What we don’t claim
Trust in this field is earned by what an organisation refuses to overstate. This ledger is here permanently, and it will be updated as facts change rather than quietly removed.
Accreditation
Not yet accredited. Assistance Dogs International accreditation is the standard VA requires for recognition under 38 CFR 17.148(c)(1), and we are not there yet. Any veteran seeking a dog today should prefer an ADI-accredited provider, and we will say so until we hold that accreditation ourselves.
Placements
We make no claim about dogs placed, veterans served, or outcomes achieved. When there are numbers, they will be published with their denominators.
The evidence
The two central trials on this page are non-randomized, drew from a single provider, and followed participants for three months. A randomized trial is underway. We describe service dogs as a promising complementary intervention, not a treatment, and not a substitute for evidence-based PTSD care.
Sleep specifically
Objective actigraphy showed no significant difference in sleep duration, onset latency, wake after sleep onset or efficiency. What improved was perceived sleep quality, fear of sleep, and nightmare frequency. We will not market a dog as something that repairs sleep architecture.
Legal information
Everything in the rights section is quoted from the primary law with citations so you can verify it. It is not legal advice, it is not a substitute for a lawyer, and regulations change — the HUD withdrawal above is a live example of exactly that.
Cost to veterans
A veteran should never be charged for a service dog by an accredited program. Raising and training a dog over roughly two years is expensive, and that cost belongs to donors and grants, not to the person who already paid.
Sources
Peer-reviewed research 14
01Pigeon WR, Campbell CE, Possemato K, Ouimette P. Longitudinal relationships of insomnia, nightmares, and PTSD severity in recent combat veterans. Journal of Psychosomatic Research 2013;75:546–50. PMID 24290044.
02Don Richardson J, King L, St Cyr K, Shnaider P, Roth ML, Ketcheson F, Balderson K, Elhai JD. Depression and the relationship between sleep disturbances, nightmares, and suicidal ideation in treatment-seeking Canadian Armed Forces members and veterans. BMC Psychiatry 2018;18:204. PMID 29921268.
03Pruiksma KE, Slavish DC, Taylor DJ, Dietch JR, Tyler H, Dolan M, Bryan AO, Bryan CJ. Nightmares and insomnia in the US National Guard: mental and physical health correlates. International Journal of Behavioral Medicine 2021;28:238–249. PMID 32495239.
04Leighton SC, Rodriguez KE, Jensen CL, MacLean EL, Davis LW, Ashbeck EL, Bedrick EJ, O’Haire ME. Service dogs for veterans and military members with posttraumatic stress disorder: a nonrandomized controlled trial. JAMA Network Open 2024;7:e2414686. doi:10.1001/jamanetworkopen.2024.14686. PMID 38833250.
05Margolies SO, Rybarczyk B, Vrana SR, Leszczyszyn DJ, Lynch J. Efficacy of a cognitive-behavioral treatment for insomnia and nightmares in Afghanistan and Iraq veterans with PTSD. Journal of Clinical Psychology 2013;69:1026–42. PMID 23629959.
06Rodriguez KE, LaFollette MR, Hediger K, Ogata N, O’Haire ME. Defining the PTSD service dog intervention: perceived importance, usage, and symptom specificity of psychiatric service dogs for military veterans. Frontiers in Psychology 2020;11. PMID 32849004.
07Jensen CL, Rodriguez KE, MacLean EL, Abdul Wahab AH, Sabbaghi A, O’Haire ME. Characterizing veteran and PTSD service dog teams: exploring potential mechanisms of symptom change and canine predictors of efficacy. PLoS ONE 2022;17. PMID 35895599.
08Bristol S, Leighton SC, Schwichtenberg AJ, Campbell RL, Ashbeck EL, Taylor DJ, Bedrick EJ, O’Haire ME. The influence of service dog partnerships on perceived and objective sleep quality for military veterans with PTSD. Frontiers in Sleep 2024. doi:10.3389/frsle.2024.1432919.
15U.S. Department of Veterans Affairs, National Center for PTSD. How Common is PTSD in Veterans? Lifetime PTSD 29% among OIF/OEF veterans; 7% among all veterans; 6% general population.
16Substance Abuse and Mental Health Services Administration, Disaster Technical Assistance Center. First Responders: Behavioral Health Concerns, Emergency Response, and Trauma. Supplemental Research Bulletin, May 2018. An estimated 30% of first responders develop behavioral health conditions including depression and PTSD, versus 20% in the general population.
17Carleton RN, Afifi TO, Turner S, Taillieu T, et al. Mental disorder symptoms among public safety personnel in Canada. Canadian Journal of Psychiatry 2018. PMID 28845686. n = 5,813; 15.1% screened positive for one disorder and 26.7% for more than one.
18Zamorski MA, Bennett RE, Rusu C, Weeks M. Prevalence of past-year mental disorders in the Canadian Armed Forces, 2002–2013. Canadian Journal of Psychiatry 2016. PMID 27270739. Past-year PTSD 5.3% in 2013; adjusted odds ratio 2.1 versus 2002.
19Medina-Mora ME, Genis-Mendoza AD, Villatoro Velázquez JA, Bustos-Gamiño M, et al. The prevalence of symptomatology and risk factors in mental health in Mexico: the 2016–17 ENCODAT cohort. International Journal of Environmental Research and Public Health 2023. PMID 36833803. n = 56,877; post-traumatic stress symptoms 5.7%.
20Meza Peña C, Bravo-Garza AP, Loredo-Villa MA. Después del enfrentamiento: estrés en policías de zona rural del noreste de México. URVIO, Revista Latinoamericana de Estudios de Seguridad 2025. doi:10.17141/urvio.43.2025.6460. n = 112 police and civil-protection personnel, rural Nuevo León; stress, anxiety, depression and burnout measured with the DASS.
United States — law and penalties 8
09Nondiscrimination on the Basis of Disability by Public Accommodations and in Commercial Facilities, 28 CFR §§ 36.104, 36.302(c); Nondiscrimination on the Basis of Disability in State and Local Government Services, 28 CFR §§ 35.104, 35.136. Text as rendered by the electronic CFR, issue date September 30, 2026.
10Nondiscrimination on the Basis of Disability in Air Travel, 14 CFR §§ 382.3, 382.72, 382.74–382.80 (Doc. No. DOT-OST-2018-0068, 85 FR 79774, Dec. 10, 2020); Transportation Services for Individuals with Disabilities, 49 CFR § 37.167; Fair Housing Act regulations, 24 CFR §§ 100.202, 100.204; VA service dog benefits, 38 CFR § 17.148 (77 FR 54381, Sept. 5, 2012).
11Puppies Assisting Wounded Servicemembers for Veterans Therapy Act, Pub. L. No. 117-37, 135 Stat. 329 (Aug. 25, 2021), §§ 2–3, adding 38 U.S.C. § 1714(e)–(f).
13U.S. Department of Housing and Urban Development, Office of Fair Housing and Equal Opportunity. Notification of Withdrawal of Fair Housing and Equal Opportunity Guidance Documents, Docket No. FR-6571-N-01, 91 Fed. Reg. 17291 (Apr. 6, 2026) (effective date of withdrawal September 17, 2025).
14Americans with Disabilities Act of 1990, 42 U.S.C. § 12101 et seq., including § 12112(b)(5)(A); Fair Housing Act, 42 U.S.C. § 3604(f)(3)(B).
3128 CFR 85.5, Civil Monetary Penalties Inflation Adjustment table (DOJ), current amounts for 42 USC 12188(b)(2)(C) (ADA Title III: $118,225 first violation; $236,451 subsequent) and 42 USC 3614(d)(1)(C) (Fair Housing Act: $131,308; $262,614). 28 CFR 36.504(a)(3).
3224 CFR 180.671, assessing civil penalties for Fair Housing Act cases before a HUD administrative law judge: $26,262; $65,653; $131,308. 42 USC 3613 (civil actions; damages and attorney’s fees).
3314 CFR 383.2 (89 FR 106292, Dec. 30, 2024), civil penalty amounts under 49 USC 46301: general $75,000 per violation; $17,062 for small businesses and individuals under 49 USC 41705.
Canada 6
21Accessible Transportation for Persons with Disabilities Regulations, SOR/2019-244, ss. 1, 51 (Canada). Text from the Justice Laws Website.
22Guide Dog and Service Dog Act, S.B.C. 2015, c. 17, ss. 1–6.1 (British Columbia). Text from BC Laws.
23Integrated Accessibility Standards, O. Reg. 191/11, ss. 80.45(4), 80.47 (Ontario), under the Accessibility for Ontarians with Disabilities Act, 2005. Text from Ontario e-Laws, currency date September 30, 2026.
34Canadian Human Rights Act, R.S.C. 1985, c. H-6, s. 53(2)(e) and (3); Canada Transportation Act, S.C. 1996, c. 10, s. 177(1)(b); Guide Dog and Service Dog Act, S.B.C. 2015, c. 17, s. 8.
37Service Dogs Act, S.A. 2007, c. S-7.5, s. 6 (Alberta): fine not exceeding $3,000 for a discriminatory practice under s. 3; $300 for falsely claiming the Act’s benefit.
38Human Rights Code, R.S.O. 1990, c. H.19, s. 45.2; Accessibility for Ontarians with Disabilities Act, 2005, S.O. 2005, c. 11, s. 37; Charte des droits et libertés de la personne, RLRQ c. C-12, s. 49. Cited by section; figures to be quoted from the statutes.
41Canadian Human Rights Act, R.S.C. 1985, c. H-6, ss. 3(1), 5, 6. Service Dogs Act, S.A. 2007, c. S-7.5, ss. 1, 3, 4, 6 (Alberta King’s Printer). Charte des droits et libertés de la personne, RLRQ c. C-12; The Service Animals Protection Act, C.C.S.M. c. S90; Service Dog Act, S.N.S. 2016, c. 4; Service Animal Act, S.N.L. 2012, c. S-13.2 — cited.
42Human Rights Code, R.S.O. 1990, c. H.19, ss. 10(1) (“disability”), 45.2(1); Accessibility for Ontarians with Disabilities Act, 2005, S.O. 2005, c. 11, s. 37(3); Blind Persons’ Rights Act, R.S.O. 1990, c. B.7, ss. 1, 2(1), 6(1). Text from Ontario e-Laws, read 2026-10-04.
México 5
24Ley General para la Inclusión de las Personas con Discapacidad, arts. 2 fr. XXVI, 16 fr. III, 17 fr. II, última reforma DOF 14-06-2024 (México).
25Ley Federal para Prevenir y Eliminar la Discriminación, art. 9 fr. XXII Quáter, adicionada DOF 04-12-2023 (México).
26Ley de Aviación Civil, arts. 47 Bis fr. I, 47 Bis 1 (México).
35Ley Federal para Prevenir y Eliminar la Discriminación, arts. 83, 83 Bis, 83 Ter (medidas administrativas y de reparación), última reforma DOF 14-11-2025.
36Código Penal Federal (México), art. 149 Ter, última reforma DOF 13-03-2026. Prison of one to three years or 150–300 days of community service, and up to 200 días multa, for denying a service on discriminatory grounds; penalty increased by one half for public servants, with removal and disqualification.
39Constitución Política de los Estados Unidos Mexicanos, art. 1°, últimas reformas DOF 02-06-2026. Ley Federal de Protección al Consumidor, arts. 58, 128 (multas actualizadas DOF 23-12-2025), 128 Bis. Ley Federal del Trabajo, arts. 3° (DOF 15-01-2026), 133 fr. I.
40Ley para la Protección de los Derechos de las Personas con Discapacidad del Estado de Nuevo León, arts. 2 fr. XXVI, 44 bis–44 bis 7 (adicionados P.O. 24-12-2008). Text from the H. Congreso de Nuevo León.
Agencies, surveys and import rules 5
12Wisch RF. Table of State Assistance Animal Laws. Animal Legal & Historical Center, Michigan State University College of Law, 2025.
27Canadian Food Inspection Agency. Import Reference Document (as referenced in the Health of Animals Regulations), dogs — personal imports; assistance-dog clause; no post-import quarantine.
28SENASICA (Servicio Nacional de Sanidad, Inocuidad y Calidad Agroalimentaria). Requirements and Procedures for Traveling to Mexico with Your Pet. gob.mx, last updated 4 August 2026. New World screwworm inspection document for dogs from the United States; inspection only from Canada.
29U.S. Centers for Disease Control and Prevention. Entry Requirements for Dogs from Dog-Rabies Free or Low-Risk Countries (reviewed 22 July 2024) and High-Risk Countries for Dog Rabies. Canada and Mexico not listed as high-risk.
30Reglamento de Autotransporte Federal y Servicios Auxiliares (México), terminal accessibility provisions; no service-animal clause.