When do Phoenix homes actually get mold?

Overwhelmingly during the monsoon — the roughly 15-week window the National Weather Service fixes at June 15 to September 30. Our analysis of federal climate and flood data found four moisture signals spike together in that window: outdoor humidity jumps about 2.5 times, home AC condensate goes from essentially zero to its annual peak, Maricopa County flood-insurance claims run about 23 times the dry-season rate, and wet materials stop drying inside the EPA's 24-to-48-hour window. The desert doesn't prevent mold — it front-loads the whole mold season into one stretch of late summer.

Does a Phoenix air conditioner really produce no condensate most of the year?

Close to it. In our psychrometric model of a typical Phoenix central AC, the system condenses about zero gallons per day for roughly nine months, October through June, because the desert air rarely reaches the cooling coil's dew point — so the AC does almost pure sensible cooling. Condensate appears almost entirely in the monsoon: about 2.1 gallons per day in July, a peak near 3.8 in August, and about 0.9 in September. Essentially 100% of a home's annual AC condensate forms in those three months. The danger isn't the volume — a humid Gulf-Coast AC makes far more — it's the timing plus the clogged-drain-line failure that sends that water into a ceiling or wall.

What share of Phoenix-area flood claims happen during the monsoon?

We pulled and aggregated every Maricopa County NFIP flood-insurance claim from 1978 to 2025 — 2,630 claims, about $26 million paid. 58% of all claims occurred in July, August and September, and 58.3% fell within the exact June 15 to September 30 monsoon window. A peak monsoon month carries about 23.5 times the claims of a typical pre-monsoon month: 565 claims in August versus an average of about 24 per month across April, May and June. Pre-monsoon accounts for just 2.7% of all claims. NFIP data covers insured flood losses only, so we use it as a sourced proxy for the timing of rain-driven water intrusion, not its absolute level.

Isn't the desert too dry for mold?

No — and the measured federal evidence is consistent on this. The EPA/HUD American Healthy Homes Survey and its ERMI mold index (Vesper et al., 2011) found water-damage-associated indoor molds are national in scope, not systematically lower in the desert Southwest; only outdoor mold groups are lower in arid regions. NHANES allergy data (Salo et al.) shows mold sensitization does not differ significantly by U.S. region. And the Institute of Medicine's 2004 report concluded essentially all of the 100-million-plus U.S. homes see excess dampness or leaks at some point. Indoor moisture, not outdoor climate, drives household mold.

If mold starts during the monsoon, does it go away when the house dries out?

No. When the wet material dries, active growth stops — but the mold goes dormant rather than disappearing. The colony and its spores survive, can still cause allergic reactions, and start growing again whenever that spot gets wet. The EPA is explicit that drying or killing mold isn't enough — it must be physically removed — and that it's impossible to clear all mold spores from indoor air; they simply don't grow without moisture. Two things also carry risk past the monsoon: dormant mold from a past event, and chronic indoor moisture — a slow plumbing or slab leak, or a recurring AC-condensate overflow — that can grow mold in any month. The roughly 15-week monsoon is when mold is most likely to start, not the only time it exists.

What should a Phoenix homeowner actually do with this?

Treat mold risk as seasonal and act before mid-June. Two moves cover most of it: have your AC condensate drain line cleared before cooling season ramps, because it's the driver most tied to how Phoenix homes run and it's the one that overflows into a ceiling; and after any monsoon water intrusion — a roof leak, a flooded room — dry the wet materials within 24 to 48 hours, the EPA's window before mold typically begins to grow. Because the risk is compressed into one window, getting ahead of that window is most of the battle.

Is this a peer-reviewed scientific study?

No, and we state that plainly. This is a transparent, reproducible analysis of public federal data — NOAA NCEI climate normals and OpenFEMA flood-claim records — combined with a standard psychrometric model built from published ASHRAE methods. It is not peer-reviewed academic research. The psychrometric figures are an order-of-magnitude model with stated assumptions and a sensitivity range, the flood data is an insured-loss proxy for timing, and the drying window is an indicator, not a predictor for any individual home. We publish the methods, the limitations, and the underlying datasets so anyone can check the work.

The Desert-Mold Paradox: When Phoenix Homes Actually Get Mold (a federal-data analysis)

A wall of monsoon dust and rain rolling at dusk across rows of single-story tile-roof stucco homes in a Phoenix valley neighborhood, with warm low sunlight catching the rooftops and desert mountains on the horizon. — The same desert that's bone-dry for nine months kicks off its entire mold season in one stretch of late summer. This analysis maps exactly when, using federal climate and flood data.

Phoenix’s mold risk is not spread evenly across the year the way it is in a humid climate. It is compressed into the roughly 15-week monsoon — June 15 to September 30 — when four things spike at once: outdoor moisture jumps about 2.5×, home AC systems go from condensing essentially zero water to their annual peak, flood-insurance claims run about 23× the dry-season rate, and wet materials stop drying inside the EPA’s 24-to-48-hour window. The desert doesn’t prevent mold. It front-loads the start of the mold season into one window — the stretch when new growth is most likely to begin. (Mold that gets going doesn’t vanish when the air dries back out; more on that below.)

In plain English

The short version. People say you can’t get mold in the desert. For about nine months a year, that’s basically true. Then the summer monsoon arrives and three things go wrong at the same time. The outdoor air turns muggy. Your air conditioner — which makes almost no water most of the year — suddenly starts producing a few gallons a day, and if its drain line is clogged, that water ends up inside a wall or ceiling. And storms push rain into roofs and rooms: 58% of the area’s flood-insurance claims over the last 47 years happened in just these three months. Wet materials also dry slower now, and mold only needs something to stay wet for 24 to 48 hours to start growing.

So the season when mold starts in Phoenix is only about 15 weeks long. That’s the good news — you can get ahead of one short window. The catch: once mold starts, it doesn’t disappear when the desert dries out. It goes dormant and waits for the next time it gets wet, and a year-round indoor leak can grow mold in any month. So the monsoon is the time to prevent mold; the rest of the year is the time to find and remove whatever already started. The federal data behind all this is below.

An original analysis by Mold Pros Phoenix, an independent Phoenix mold information resource — built entirely on public federal data. The methods, limitations, and underlying datasets are published in full below, so anyone can check the work.

The paradox

“You can’t get mold in the desert” is one of the most repeated beliefs about Phoenix homes, and it is wrong — but the usual rebuttal is also incomplete. Most corrections simply note that mold is a moisture problem rather than a humidity problem and leave it there. True, but it skips the more useful question: if desert homes get mold, when do they get it, and why then?

This analysis answers that question with federal data. It does three things no templated “mold in the desert” page does. First, it builds an original psychrometric model of what a Phoenix air conditioner actually does to water across the year. Second, it pulls and aggregates every Maricopa County federal flood-insurance claim back to 1978 to show the real seasonality of rain-driven water intrusion. Third, it lines those up against published federal health-and-housing surveys to confirm the desert is not mold-immune. The result is a single, specific finding: in Phoenix, the season when mold gets started is not the year — it is the monsoon.

Finding 1 — The hidden water: what a Phoenix AC actually does

Start where most Phoenix homeowners never look — inside their own air conditioner.

In a humid climate, an AC is a dehumidifier as much as a cooler: it runs cold coils that pull liquid water out of the air all summer, which is why a Houston or New Orleans system drips a steady stream of condensate. In Phoenix, for most of the year, that barely happens. We modeled a typical Phoenix central AC using a standard steady-state moisture balance (the method is in the Methodology below), and the result is striking: the system condenses about 0 gallons per day for roughly nine months, October through June. The desert air is so dry that it rarely reaches the cooling coil’s dew point, so the AC does almost pure sensible cooling — it drops the temperature without wringing out water. That matches HVAC field experience, where dry-climate units are known to produce very little condensate.

Then the monsoon arrives and the picture flips. Condensate appears almost entirely in three months: about 2.1 gal/day in July, a peak of about 3.8 gal/day in August, and about 0.9 gal/day in September. Put another way, essentially 100% of a home’s annual AC condensate forms in July through September. Across a full year the model puts a typical home at about 210 gallons (a sensitivity range of 155–258 gallons across realistic infiltration and internal-moisture assumptions — and the monsoon-concentration finding holds in every case).

Bar chart of modeled AC condensate in gallons per day for a typical Phoenix central air conditioner by month. The value is about zero from October through June, then rises to about 2.1 in July, peaks near 3.8 in August, and falls to about 0.9 in September. — A Phoenix AC condenses almost no water for nine months, then produces essentially its entire annual condensate in the monsoon. The takeaway: the AC moisture problem here is seasonal, not constant.

Scale that to the metro and it becomes a genuinely large hidden water source. With roughly 1.62 million air-conditioned homes in the Valley (1.88 million MSA households, about 86% with central AC), the model implies about 340 million gallons of AC condensate per year across Phoenix — essentially all of it during the monsoon.

Here is the part that matters for mold, and it cuts against the obvious read. Even at its August peak of about 3.8 gal/day, a Phoenix AC produces far less condensate than a humid Gulf-Coast unit, which can run 10–20 gal/day. So the Phoenix mold danger from air conditioning is not the volume of water — it’s the timing plus the failure mode. A condensate drain line that clogs with dust and algae, or a rusted drain pan, sends that monsoon-season water somewhere it shouldn’t go: into a ceiling cavity, down a wall, across an attic floor. Out of sight, it can sit there past the 24-to-48-hour window and grow mold. The water volume is small; the consequence of misdirecting it is not.

The reason the timing lines up so tightly with the rest of this report is the outdoor air itself. The humidity ratio — the actual mass of water in the outdoor air — jumps about 2.5× from May to August. Mean dew point roughly triples, from around 34 °F in the dry months to about 58 °F in August. That is the same moisture surge that loads the cooling coil, and, as the next finding shows, it is the same surge that puts water on roofs and into walls.

Finding 2 — When the Valley actually floods

If AC condensate is the moisture source hiding inside the house, rain-driven intrusion is the one hiding in plain sight — and to measure its real seasonality, we went to the most complete public record of Valley water damage that exists: federal flood-insurance claims.

We pulled and aggregated every Maricopa County claim ourselves from the OpenFEMA database. The dataset is 2,630 Maricopa County NFIP flood-insurance claims from 1978 to 2025, totaling about $26 million paid. The seasonality is not subtle. 58% of all claims occurred in July, August and September, and 58.3% fell within the exact June 15 to September 30 monsoon window. A single peak monsoon month carries about 23.5× the claims of a typical pre-monsoon month — 565 claims in August against an average of roughly 24 per month across April, May and June. The pre-monsoon stretch (April plus May plus June combined) accounts for just 2.7% of all claims. And 53% of claims are coded “accumulation of rainfall or snowmelt” — i.e., water that built up faster than it could drain, exactly the kind of event that soaks building materials.

Bar chart of Maricopa County federal flood-insurance claims by month from 1978 to 2025. July, August and September tower over every other month at 522, 565 and 439 claims, together 58 percent of the total, while April, May and June are nearly empty at 31, 30 and 11. A smaller secondary bump appears in February at 290. — Maricopa County flood-insurance claims cluster hard in July–September. A peak monsoon month runs about 23 times a typical pre-monsoon month — the seasonality of rain-driven water intrusion in one chart.

Three honest nuances belong with this data. First, NFIP records are insured flood losses only, so they undercount total water intrusion — most monsoon roof leaks and street-flooding events never become a federal flood claim. We are not using this as a count of how much water gets into Phoenix homes; we are using it as a well-sourced, decades-deep proxy for the timing and seasonality of rain-driven intrusion, which is the upstream driver of post-flood mold. Second, Phoenix has a real secondary winter wet season — you can see the smaller February bump (290 claims) in the chart, driven by Pacific frontal storms in late winter and autumn — but the monsoon dominates by a wide margin. Third, the dates and ZIP codes are insurer-reported, with the ordinary imperfections that implies. None of those caveats move the headline: Valley flooding is a summer phenomenon, and so is the water intrusion that precedes mold.

The convergence

Put the two original findings next to the climate record and the centerpiece of this report appears: four independent moisture signals all spike in the same 15-week window. They are not loosely correlated. They peak in the same month.

Monthly chart overlaying three normalized series for Phoenix that all spike together in July through September inside a shaded monsoon window: mean dew point in degrees Fahrenheit, AC condensate in gallons per day, and FEMA flood claims by count. Each series is flat and low from January through June, then climbs sharply and peaks in August before falling back in autumn. — The monsoon convergence — three different moisture signals, three different units, one shared peak in August. This single-window concentration is what 'the desert-mold paradox' actually means.

The four-way convergence, stated plainly:

Outdoor moisture jumps about 2.5× from May to August (mean dew point roughly triples to about 58 °F), wetting roofs, walls, and the air the AC has to cool.
AC condensate goes from essentially zero to its annual peak of about 3.8 gal/day in August — the home’s own hidden water source switching on.
Flood-insurance claims run about 23× the dry-season rate, with 58% of all Maricopa claims landing in July–September.
Wet materials stop drying inside the EPA’s window. When humidity is at its annual high, materials that get wet — from a roof leak, a condensate overflow, a flooded room — are far less able to dry within the 24-to-48-hour period the EPA identifies as the threshold before mold typically begins to grow.

Any one of these alone would raise mold risk. The desert-mold paradox is that all four arrive together, in one compressed window, in a climate everyone assumes is too dry for mold. That convergence — not the annual average humidity, which is genuinely low — is what makes the late-summer Valley a mold environment for the homes where water gets loose.

Finding 3 — The desert is not mold-immune

The convergence explains when. But it is worth confronting the underlying myth directly with measured evidence, because the strongest version of “you can’t get mold in the desert” claims that desert homes simply have less mold, period. The published federal record says otherwise. We did not re-analyze this restricted microdata ourselves — these are published findings from federal surveys — but they all point the same way.

EPA / HUD American Healthy Homes Survey + ERMI. The Environmental Relative Moldiness Index work by Vesper et al. (2011) found that water-damage-associated indoor molds are “national in scope” — not systematically lower in the desert Southwest. Only the outdoor mold groups are lower in arid regions. The indoor mold that affects homes tracks indoor moisture, not the dryness of the air outside.
NHANES. National allergy-sensitization data analyzed by Salo et al. found that mold and Alternaria allergy sensitization does not differ significantly by U.S. census region. People in dry regions are sensitized at broadly the same rates as people in humid ones.
Institute of Medicine / National Academies (2004). The CDC-sponsored report Damp Indoor Spaces and Health concluded that essentially all of the 100-million-plus U.S. homes experience excess dampness or leaks at some point, and that dampness and mold are linked to respiratory effects such as coughing, wheezing, and worsening of asthma.

Read together, the measured evidence and our climate-and-flood analysis tell one consistent story. Outdoor desert air is dry; indoor moisture is not governed by it. Household mold is driven by water that gets into the building — and in Phoenix, the data shows that water arrives on a schedule.

Does the mold go away when the desert dries out?

Short answer: no — and because the “15-week” framing can hide it, it’s worth stating plainly. The monsoon is when mold is most likely to start. It is not a window mold politely confines itself to. Two things carry the risk into the other nine months.

The science here is the EPA’s, not ours. The agency states flatly that “it is impossible to get rid of all mold and mold spores indoors” — spores are always present, and they simply “will not grow if moisture is not present.” That cuts both ways: dry a surface out and growth stops, but “dead mold may still cause allergic reactions,” so “it is not enough to simply kill the mold, it must also be removed.” Drying is not remediation. (Source: EPA, A Brief Guide to Mold, Moisture and Your Home.)

So read this report the right way. The ~15 weeks is the ignition season — the time to get ahead of new growth. The rest of the year is when you find and remove anything that already started, and stay alert to the indoor leaks that don’t wait for the monsoon. A concentrated risk is easier to prevent; it does not make existing mold disappear.

What it means for a Phoenix homeowner

The practical upshot is genuinely good news, because a concentrated risk is a manageable one. To stop new mold, you are not fighting the whole year — you are getting ahead of one window. (The exceptions, from the section above: mold that already started, and year-round indoor leaks. Those don’t wait for the monsoon.)

Clear the AC condensate drain line before the monsoon. This is the driver most tied to how Phoenix homes actually run, and the line that overflows into a ceiling or wall is the failure mode that turns a small amount of water into a full mold remediation job. Have it cleared before mid-June, when condensate switches on.
Dry any water intrusion within 24 to 48 hours. After a monsoon roof leak or a flooded room, the EPA’s window is short but real: materials dried inside it usually don’t grow mold. During the monsoon, with humidity at its annual peak, materials dry slower — so move faster.
Treat the risk as seasonal, and act before mid-June. The single most useful thing this analysis offers a homeowner is timing. The work to do — checking the roof and flashing, clearing the condensate line, grading water away from the foundation — should happen in spring, before the four signals converge.

If you want the room-by-room and driver-by-driver picture behind this timing, our companion 2026 Phoenix Mold Risk Report maps the seasonal calendar and the five physical moisture drivers, and Phoenix mold statistics covers how common reported mold actually is across the metro. The myth itself gets the full treatment in mold in the desert. If you’ve already found mold or can smell it but can’t see it, the right first step is to confirm the moisture source, which is what a professional mold inspection is built for. The full library of Phoenix mold guides covers each growth type and room in depth.

Methodology

This report combines two original analyses of public federal data with a review of published federal findings. Everything here is reproducible from the cited sources and the datasets linked below.

Psychrometric model (Finding 1). Humidity ratios were computed per the ASHRAE Handbook—Fundamentals, Chapter 1, using W = 0.621945·pw/(P − pw), with saturation vapor pressure from the Arden Buck (1996) equation. Climate inputs are the NOAA NCEI 1991–2020 Climate Normals for Phoenix Sky Harbor (station GHCND:USW00023183), with monthly mean dew point derived from the NCEI hourly normals; barometric pressure is 97.6 kPa at the station’s 1,086 ft elevation. AC condensate is the difference between incoming and setpoint moisture across a steady-state home moisture balance: internal latent generation (~20 lb/day per ANSI/ASHRAE Standard 160-2009b) plus infiltration (~0.4 ACH) against an indoor setpoint of 75 °F / 50% RH, with a coil apparatus dew point of ~52 °F. The annual sensitivity range (155–258 gallons) spans infiltration of 0.25–0.6 ACH and internal moisture of 15–24 lb/day. Metro scaling uses ~1.88 million MSA households at ~86% central-AC prevalence (EIA RECS 2020, regional).

Flood-claim aggregation (Finding 2). We queried the OpenFEMA “FIMA NFIP Redacted Claims v2” public API for all Maricopa County, Arizona claims (1978–2025), then aggregated by month of loss. Monsoon-window figures use the National Weather Service Phoenix standard of June 15 – September 30. NFIP covers insured flood losses only and is used here as a proxy for the timing of rain-driven water intrusion, not its absolute magnitude.

Measured-evidence review (Finding 3). The American Healthy Homes Survey / ERMI, NHANES, and Institute of Medicine findings are published federal results, cited in place. We did not re-analyze that restricted microdata ourselves.

Downloadable data

The underlying datasets are public so anyone can check or extend the analysis:

AC condensate by month: phoenix-ac-condensate-by-month.csv — monthly temperature, dew point, humidity ratio, and modeled condensate.
Maricopa flood claims by month: maricopa-nfip-claims-by-month.csv — NFIP claim counts and share of dated claims by month.

Suggested citation

Mold Pros Phoenix (2026). The Desert-Mold Paradox: When Phoenix Homes Actually Get Mold. moldprosphoenix.com/guides/desert-mold-paradox/

Limitations

Transparency is the point of this report, so its limits are stated as plainly as its findings.

The psychrometric model is order-of-magnitude, not a measurement. It is a standard steady-state model built from published ASHRAE methods and NOAA normals, with explicit assumptions and a stated sensitivity range — not metered data from real homes. The headline finding it supports is the monsoon concentration of condensate, which holds across the full sensitivity range; treat the specific gallon figures as modeled estimates.
Flood claims are an insured-loss proxy. NFIP data captures only insured flood losses and undercounts total water intrusion. It is used for the seasonality of rain-driven intrusion, which it documents well over nearly five decades, not for the absolute level of water damage. Dates and ZIPs are insurer-reported.
The drying window is an indicator, not a predictor. The EPA’s 24-to-48-hour threshold describes when materials generally begin to grow mold; it does not predict whether any specific home will. Real outcomes depend on the material, the temperature, and how completely it dried.
The measured-evidence findings are cited, not re-derived. We did not re-analyze the AHHS/ERMI, NHANES, or IOM microdata; we report their published conclusions.
This is not peer-reviewed academic research. It is a transparent, reproducible analysis of public federal data by an independent Phoenix mold information resource, published with its methods, limitations, and datasets so it can be checked. It is not a clinical or engineering study, and nothing here is medical advice; for health questions about mold exposure, consult the CDC and a medical professional.

Sources

NOAA NCEI — 1991–2020 U.S. Climate Normals, Phoenix Sky Harbor (GHCND:USW00023183): monthly temperature and dew point inputs. U.S. Climate Normals
OpenFEMA — FIMA NFIP Redacted Claims v2: source for all 2,630 Maricopa County flood-claim records (1978–2025). FIMA NFIP Redacted Claims v2
ASHRAE Handbook—Fundamentals — psychrometric methods (humidity ratio, Ch. 1). ASHRAE Handbook
Arden Buck (1996) — saturation vapor pressure equation used for dew-point conversions. Arden Buck equation
ANSI/ASHRAE Standard 160 — moisture-design loads (internal latent generation). Standards and Guidelines
U.S. EPA — mold cleanup and the 24-to-48-hour drying window. Mold Cleanup in Your Home
EPA / HUD — Vesper et al. (2011) — ERMI / American Healthy Homes Survey: water-damage molds are national in scope. ERMI study (PMC)
Institute of Medicine / National Academies (2004) — Damp Indoor Spaces and Health. National Academies
U.S. EIA — Residential Energy Consumption Survey (RECS) 2020: central-AC prevalence for metro scaling. RECS
NWS Phoenix — Arizona monsoon defined as June 15 – September 30. National Weather Service Phoenix