Beyond the 36.

Why ERMI selected the 36

The Environmental Relative Moldiness Index (ERMI) was developed by EPA researchers in the mid-2000s based on the American Healthy Homes Survey (AHHS). The 36 species were chosen for their statistical association with moisture-damaged versus reference homes in a population-scale survey of U.S. housing stock,not for completeness as an indoor mycology inventory. The panel splits into 26 water-damage indicator species (Group 1) and 10 environmental reference species (Group 2); the ERMI score subtracts the sum of log-concentrations for Group 2 from the sum for Group 1.

The fundamental EPA framing has always been that ERMI was developed for population-level research, not individual home diagnostic certainty. The 36-species selection optimizes for statistical separation between water-damaged and reference homes at the population level. Individual-home interpretation is a downstream application, useful, but with caveats the original population-level method does not address.

The EPA's original MSQPCR panel actually tested 82 species or assay groups. The Vesper et al. 2007 ERMI was developed from this panel using a 36-species selection, 26 Group 1 (water-damage indicator) plus 10 Group 2 (environmental reference). A related earlier paper (Meklin et al. 2004) applied a stricter threshold of at least 2 spores per 5 mg dust (≈ 0.4 SE/mg) which retained 25 species/groups for moldy-vs-reference analysis. See the 82-species story for the methodology behind both selections. That page is the anchor for how to read any extended-panel result.

The mycotoxin-producer gap

The most consequential gap in the standard 36 is three major mycotoxin families with no producer organism on the panel:

• Fumonisins, produced by Fusarium verticillioides and F. proliferatum; no Fusarium species on the standard ERMI.
• Trichothecenes from Fusarium (DON, T-2, HT-2, NIV); same producer gap.
• Zearalenone, produced by F. graminearum and F. culmorum; same producer gap.

The Fusarium omission is the single most consequential one. Adding a Fusarium target (genus or species level) addresses the producer-organism route for fumonisins, zearalenone, and the Fusarium-family trichothecenes simultaneously.

Other producer gaps with extended-panel coverage:

The water-damage indicator gap

Several species are strong indicators of active building moisture damage but are not on the EPA's 36-species panel. Two with current extended-panel coverage:

The Ulocladium genus in general is among the more reliable water-damage indicator groups in indoor mycology, the absence of a Ulocladium target from the standard panel is one of the more puzzling historical choices in the EPA's 36-species selection.

What extended panels typically don't add

Some lab extended panels include species whose presence is not actionable in a homeowner-scale building assessment:

• Clinical pathogens (Sporothrix, Histoplasma, Cryptococcus, Pseudallescheria), relevant to hospital surveillance, not residential building-level investigation.
• Yeasts (Candida albicans, Candida auris), clinical rather than building-coded detection.
• Allergen-only species (some Bipolaris and Curvularia), relevant for clinical allergy testing, not for assessing building moisture.

Important disclosure

Extended species do not have national reference distributions like the EPA AHHS dataset used for ERMI. Interpretation of extended-species results is qualitative, based on indoor mycology literature. No lab's extended interpretation should be confused with EPA-validated ERMI scoring; modified or “extended” ERMI scores produced by labs are not EPA-validated calculations.

Further reading

• The 82-species story, methodology anchor
• All 18 extended species, index
• Penicillium species overview
• Indoor-relevant mycotoxins
• ERMI limitations: what dust testing does and doesn't measure