How Common Is Separation Anxiety in Dogs, Really? What the Studies Actually Show (2026)

1So How Common Is It, Really?

If you want a rough sense of scale rather than a diagnosis: using the 17.2% figure against roughly 63 million US dog-owning households, secondary sources arrive at an estimate of about 10.8 million affected dogs nationwide. Treat that as a back-of-envelope calculation, not an independent count — there is no national census of diagnosed separation anxiety, so any “X million dogs” figure you see online is derived the same way, not independently measured.

2Same Data, Two Headlines: Why One Study Said 85.9% and Another Said 9.3%

This isn’t a story about one team being wrong and one being right about dogs — both used the same real survey answers from the same 43,517 dogs. The disagreement is entirely about how to convert a 0–4 owner-rating scale into a yes/no “problem” label. Serpell and Powell’s broader critique covered more than separation anxiety: the original analysis also reported 55.6% of dogs with moderate-to-severe aggression and 49.9% with moderate-to-severe general anxiety, and the re-assessment brought those down to a 1.2–30.4% range for aggression types and 14.1–28.2% for anxiety/fear types. It’s the same pattern behind why dog food allergy prevalence estimates range from 0.2% to 40% depending on diagnostic criteria — a wide published range is usually a measurement-definition story, not a sign that nobody knows anything.

3Does the Diagnostic Bar Matter That Much? The 5% Study

This is the same mechanism driving the 85.9%-vs-9.3% gap in the previous section, just demonstrated on a different, smaller dataset and stated directly by the study’s own authors rather than inferred from a later re-assessment. It’s a useful sanity check for owners trying to interpret conflicting training advice, too — the same definitional looseness is part of why training-method “success rate” claims often don’t hold up under closer research: a loose definition of success inflates the number the same way a loose definition of “separation anxiety” does.

4Did COVID Cause a Separation Anxiety Epidemic? The Data Says No

UK LONGITUDINAL SURVEY, n=1,807 DOGS (PMC8868415, 2022)

22.1%
20.2%
17.2%
Feb 2020 (pre-lockdown)
May–Jun 2020 (lockdown)
Oct 2020 (post-lockdown)

This matters because it’s a methodology lesson as much as a pandemic-trivia fact: this is owner-reported data, not clinical diagnosis, which the study’s own authors flag as a limitation. But it’s still the best available before/after comparison on a fixed group of dogs, and it directly contradicts a claim repeated constantly in casual pet-media coverage. It’s worth separating this from a related but different question: how a dog’s routine attachment and bonding habits — like whether a dog sleeps on the bed or couch — interact with being left alone is simply not what this dataset measured. Owners sometimes also confuse ordinary forgetfulness for anxiety; the research on how a dog’s memory actually works is a useful companion read if you’re trying to rule out cognitive explanations for a behavior change.

5Are Shelter Dogs Really at Higher Risk?

Recently adopted and rehomed dogs are, functionally, a different population from dogs who have lived in one home for years — which is one more concrete reason raw prevalence numbers quoted online disagree so sharply depending on which study they’re pulled from. It also connects to a broader pattern worth knowing before adopting: unresolved behavior problems are a documented driver of dogs being returned to shelters, and separation-related distress in the first six months is exactly the kind of problem that, left unaddressed, can contribute to a return. It can also ripple into conflict in multi-dog households when a newly adopted dog’s distress at being left alone affects resident dogs too.

7Risk Factors the Data Actually Supports

If your dog is already noise-sensitive — startling at thunderstorms, fireworks, or garbage trucks — that alone is a documented reason to watch more closely for separation-related signs, not a coincidence. This is also exactly the kind of risk profile worth weighing before committing to a schedule that leaves a dog alone for long stretches; it’s part of why matching a dog to a full-time work schedule is as much about temperament risk factors as it is about breed energy level. Early in a new dog’s life, some clinginess is developmentally normal rather than pathological — the first-month adjustment period covers how to tell ordinary settling-in behavior apart from a genuine emerging pattern. When management or training is warranted, the same evidence base that applies broadly to dog training favors reward-based methods over aversive correction, particularly for anxiety-driven behavior rather than simple disobedience.