Methodology
Last updated: September 2026
Overview
The RetireHere Score is a composite rating, displayed on a 0–100 scale, summarizing community sentiment about a city across seven dimensions: Cost of Living, Healthcare Access, Cultural & Social Scene, Weather Experience, Safety Experience, Downtown Charm & Walkability, and Senior Amenities. It is built from individual reviewer ratings and appears once a city has accumulated at least 10 qualifying reviews, so a score is never shown when there isn't yet enough community input behind it.
Data Sources
RetireHere combines publicly available government and industry data with ratings submitted directly by our community of users. Our data covers natural disaster risk, climate, elevation, population, home values, and healthcare access, drawn from established public and industry sources.
Dimension Ratings
Each reviewer rates a city across seven dimensions on a 1–5 scale, or may mark a dimension "No Opinion" if they do not feel qualified to answer it. For example, a short-term visitor may not feel able to speak to healthcare access or senior amenities the way a resident could. A "No Opinion" response is excluded from that dimension's calculation; it is never treated as a neutral or average rating. Each dimension's score is built entirely from the reviewers who answered it, then the seven dimension scores are averaged together to form the composite.
The 0–100 Scale
The underlying 1–5 star composite is converted to a 0–100 score using a fixed, published conversion rather than a straight linear scale. A city where every dimension rates as simply "fine" (3 out of 5) lands in the low 60s, a passing and unremarkable score, rather than at 50, which would misrepresent neutral sentiment as a failing grade. Genuinely excellent cities reach the mid-80s to low-90s; genuinely poor cities fall into the 30s or below. A perfect 100 is not attainable by design.
Weighting for Sample Size
A RetireHere Score doesn't display until a city reaches 10 qualifying reviews. Below that threshold, the city shows as "Pending" rather than presenting a score built on too small a sample. Once a city clears 10 reviews, each of the seven dimensions still starts from a neutral baseline that real reviews progressively outweigh as more accumulate, so a single review, whether positive or negative, cannot swing a city to the extreme ends of the scale. Scores become more precise, and more confident, as review volume grows beyond that minimum. We display the number of reviews behind every score so users can judge that confidence for themselves.
Safety as a Guardrail
Safety Experience is treated differently from the other six dimensions. Strong ratings elsewhere cannot fully offset a poor Safety rating in the composite score. A city with excellent amenities but real safety concerns will not be presented as an unqualified high score. This reflects that safety is not simply one preference among many for a retirement destination, but a baseline concern most users weigh above the rest.
Handling Missing or Unavailable Data
RetireHere's underlying data is drawn from multiple public sources, and not every data point is available for every city. Rather than use different wording in different places for what is fundamentally the same handful of situations, we use four consistent terms across the entire site:
- Not Available. The source simply does not have a usable value for this city. This is the most common case, and it is never counted against a city when filtering.
- Not Applicable. The measurement genuinely does not apply to this location, which is a different fact from the data being missing. Coastal flood risk for a landlocked city is the clearest example: there is no value to report because the hazard itself cannot occur there.
- Not Rated. The thing being measured is identified and exists, but the authority that would rate it has not published a rating for it. Currently this applies only to hospital quality; see below for exactly what it covers.
- Pending. This term is used only for the RetireHere Score itself, when a city has not yet reached the 10 review minimum described above.
Two specific cases worth calling out:
- Disaster risk ratings marked "Not Applicable," "Not Available," or otherwise missing are never counted against a city when filtering. A city is not excluded from search results simply because a particular hazard does not apply to it, or because that data point was not available.
- Population values of exactly zero are displayed as "Not Available" rather than "0." This typically reflects small or remote communities with no permanently registered residents in Census data, rather than an error.
Healthcare Access
Healthcare Access is measured two ways on RetireHere, deliberately kept separate. The Healthcare Access dimension reflects reviewer sentiment. It captures how residents feel their care has been. The Healthcare Access filters described below are objective, sourced facts: how far the nearest hospital is, how that hospital is rated, and how available primary care is nearby. A dataset can't tell you how a hospital treated you, and a review can't tell you how many miles away it is. We treat these as two different kinds of information rather than blending them into one number.
The filters cover three separate questions on purpose, rather than one combined score:
- Distance to Nearest Hospital. This is the straight-line distance from the city to the nearest qualifying hospital, not driving distance or driving time. The qualifying set is limited to hospitals HIFLD lists as open and classifies as General Acute Care or Critical Access; specialty facilities such as psychiatric or rehabilitation hospitals are excluded, since a nearby specialty facility with no emergency department would be a misleading answer to "how far is the nearest hospital."
- Hospital Quality Nearby. This is the CMS Overall Star Rating of that same nearest hospital, not the best-rated hospital within some radius. A five-star hospital thirty miles away and a two-star hospital two miles away are different situations, and we show them as two separate facts rather than one blended figure.
- Primary Care Doctor Access. This is a county-level estimate of residents per primary care physician. It reflects the surrounding county, not the city specifically, so two towns in the same county will show the same figure. The category thresholds (Strong, Adequate, Limited, Severe Shortage) are anchored to HRSA's own federal shortage-designation criteria, not set arbitrarily.
A hospital shows as "Not Rated" for two different underlying reasons, and we deliberately merge them into that single label rather than distinguishing them for the user. The first reason: CMS reviewed the hospital and found insufficient data to publish a star rating, which is common for small and critical access hospitals. The second: our own matching process was not able to confidently connect the hospital's location record to its CMS record, so no rating is available to show even though CMS may have actually published one. In both cases a specific, named hospital has been identified for that city; the gap is only in the rating, never in whether a hospital was found nearby. We merge these two cases because the distinction is about which of our data sources knows about the hospital, not about the hospital itself, and it is not a distinction that should change a user's search decision. An unrated hospital is included by default rather than treated as low quality; this can be changed for any individual search.
Two limitations worth stating plainly. CMS matching coverage is measurably weaker in remote Alaska, so a Not Rated result is more likely there than elsewhere. And hospital address data is not reliably available for every listed hospital. Confirmed directly against the source data, some records return a facility's name in the address field rather than a street address. Where this occurs, no address is shown rather than displaying incorrect or duplicated information.
A Note on Data Completeness
RetireHere's data is drawn from large public datasets covering more than 32,000 U.S. cities and towns. As with any dataset of this size, occasional gaps or inconsistencies exist. We investigate and correct these where we can, and we would rather disclose our limitations plainly than obscure them.
One specific example: filtering by Climate criteria excludes 18 cities and towns for which reliable climate station data was not available at the time of publication. These places remain fully visible and searchable outside of climate-based filtering. The gap is in that one data source, not in the city's listing. As with all RetireHere data, our underlying figures are drawn from established public and industry sources rather than any editorial decision to omit a location.
My Favorite Cities
84.9°F 
$268k