California’s Pesticide Use Reporting (PUR) system records every commercial pesticide application in the state at roughly one square mile of resolution, with a calendar date, the active ingredient, and the reported mass. No other state in the country has this. The federal alternative, the USGS Pesticide National Synthesis Project, reports estimated annual pounds at the county level. The difference in resolution is large in two dimensions at once: a single California county spans hundreds to thousands of these square-mile sections, and PUR resolves applications to the day rather than to the year.
For Community Health Needs Assessments in California, this changes what is analytically possible. A CHNA prepared in Fresno County can describe specific square-mile sections within the county where pesticide application is heaviest, name the months when applications cluster, identify the specific compounds being applied, and place that exposure context next to the clinic catchments, school locations, and residential ZIP codes of the population served. A CHNA prepared in Iowa cannot do any of this. The data does not exist.
The honest case for California CHNA practitioners turns out to be subtler than “show me where pesticides cause asthma.” It is the resolution moat itself, paired with the peer-reviewed exposure-response literature that has already done this work correctly in a different California cohort. The rest of what follows is in service of that case.
What “section-level” actually means
The Public Land Survey System (PLSS) divides the western United States into a grid of townships, ranges, and sections. A section is approximately one square mile (640 acres). California’s PUR records each commercial pesticide application to the PLSS section, with a date, the EPA-registered active ingredient, and the reported pounds applied. Restricted-use materials carry additional permit data.
For 2022, the most recent year ingested into the Banana Analytics platform, this produced records for 22,503 distinct sections across California, totaling roughly 151 million pounds of active ingredient applied. (The platform’s statewide total runs below CDPR’s own published figure for the same year, roughly 181 million pounds for 2022, because the platform applies QA filters that drop error-flagged and unparseable records while CDPR applies outlier adjustment instead. For reference, CDPR’s published total for 2023 was 176 million pounds.) Across 858 distinct active ingredients. Across 12 monthly increments per year.
This is the structural fact that distinguishes California’s exposure data from anything else in the country.
Fresno County across 2020 to 2022
Fresno is the largest agricultural producer in California by farm gate value and one of the most pesticide-intensive counties in the state. The platform’s 2020–2022 window for Fresno shows:
- 81.1 million pounds of active ingredient applied across the three years, an average of roughly 27 million pounds per year
- 860,742 distinct applications recorded
- Application activity in 2,264 distinct PLSS sections (the rough equivalent of 2,264 square miles, though sections do not nest into administrative geographies cleanly)
- The mass applied declined about 29% from 2020 to 2022, consistent with the statewide downward trend CDPR reports
The most applied compound in Fresno County across the window was sulfur (24.5 million pounds), a fungicide and acaricide widely used in grape, citrus, and tree fruit production. Sulfur is not on California’s Proposition 65 list and is not a Toxic Air Contaminant, but airborne sulfur dust is a documented respiratory irritant in the EPA’s own Sulfur Reregistration Eligibility Decision. The peer-reviewed literature on sulfur and pediatric lung function is discussed in the next section.
The next most applied compounds were mineral oil (23.1 million pounds), used as a horticultural spray oil and adjuvant, and metam-potassium (6.5 million pounds), a soil fumigant that breaks down to methyl isothiocyanate (MITC), which is a CDPR-designated Toxic Air Contaminant. Telone (1,3-dichloropropene), the soil fumigant subject to recent CDPR regulatory action under Regulation 22-005 (effective January 1, 2024) and DPR 24-001 (effective January 1, 2026), accounted for 4.2 million pounds. Both glyphosate salts, listed on California’s Proposition 65 for cancer (effective 2017) though with the warning requirement permanently enjoined under a 2023 Ninth Circuit ruling, accounted for approximately 5.1 million pounds combined.
The compound mix matters because the fumigants and the dust-mobile compounds carry different exposure pathways than the herbicides. Fumigants generate air emissions hours to days after application. Sulfur drifts as dust during dry application. Glyphosate is detected in California home dust samples at rates above 98% in recent studies. These are not equivalent exposure pathways and a CHNA that treats “pesticides” as a single category misses what the data actually shows.
The most intensive single square mile
The most pesticide-intensive single section in Fresno County over 2020 to 2022 was T15S R17E Section 26 (Mount Diablo Meridian), with 375,357 pounds of active ingredient applied across 509 distinct applications and 68 distinct active ingredients. The most frequently treated section, T14S R18E Section 31, received 5,235 separate applications across the three years, which averages to more than 30 applications per week across the entire 156-week period. The total mass on T14S R18E Section 31 was 357,589 pounds. The top compounds on that section were mineral oil (561 applications), peroxyacetic acid (357 applications), hydrogen peroxide (357 applications), and the insecticides spinetoram and methoxyfenozide. The chemical mix is characteristic of a high-value perennial crop block under frequent low-dose treatment, likely table grape or tree fruit. (The PUR data does not carry the crop or site code at the silver-table layer; the crop is inferred from the chemical mix.)
The point is not that these particular sections are uniquely hazardous. The point is that within Fresno County, application intensity varies by three orders of magnitude across sections. A CHNA that uses Fresno’s county-aggregate pesticide figure as a community exposure metric is averaging the most intensive square mile in the county with hundreds of sections that received no applications. That average tells the CHNA practitioner very little about what specific populations in specific places are actually exposed to.
What the peer-reviewed literature has found
The peer-reviewed work on California pesticide exposure and respiratory health has used PUR data correctly: with residential proximity modeling, real (not modeled) health outcomes, and confounder control. The CHAMACOS cohort (UC Berkeley, Salinas Valley) is the most rigorous longitudinal study and has produced several findings relevant to Fresno’s compound mix.
Raanan and colleagues (2017, Environmental Health Perspectives) found that in CHAMACOS children, each tenfold increase in elemental sulfur applications within 1 kilometer of a child’s residence in the prior year was associated with lower lung function (FEV1), more asthma symptoms, and approximately 3.5 times higher odds of asthma-medication use. Given that sulfur is Fresno County’s most-applied compound and that elemental sulfur drift is a well-documented respiratory irritant, this is the single most operationally relevant citation for Fresno CHNA work. The Salinas Valley cohort is methodologically transferable to the San Joaquin Valley.
Raanan and colleagues (2015, Environmental Health Perspectives) found that prenatal and early-childhood urinary organophosphate metabolites were associated with respiratory symptoms and exercise-induced cough at ages 5 to 7 in the same cohort. Organophosphates are less prominent in Fresno’s current compound mix than they were a decade ago (California’s chlorpyrifos cancellation took effect at the end of 2020), but the older birth cohorts in CHNA-relevant populations carry the prior exposure history.
Gunier and colleagues (2018, Environmental Research) tested residential proximity to agricultural soil fumigants (methyl bromide, chloropicrin, metam-sodium, 1,3-dichloropropene) and respiratory health at age 7 in CHAMACOS. They found no adverse association. This is a null result and we include it here because intellectual honesty requires citing the studies that cut against a simple fumigant-asthma narrative as well as the ones that support it. The exposure-response evidence for fumigants and pediatric respiratory outcomes in California is mixed; a CHNA that asserts a clean causal story about fumigants and asthma is overclaiming.
Madrigal and colleagues (2024, Environment International) quantified the residential drift pathway directly: in carpet dust samples from 578 California children’s homes across 35 counties, applications within approximately 4 kilometers of a home predicted indoor concentrations of agricultural compounds. Glyphosate was detected in 98.6% of home samples. The Madrigal study does not measure respiratory outcomes; it quantifies the exposure pathway from agricultural application to indoor residential contamination.
The Salam and colleagues study from 2004 in the Children’s Health Study (USC, Southern California) found large odds ratios for asthma onset among children with first-year-of-life pesticide and herbicide exposure (OR ~2.4 for pesticides, ~4.6 for herbicides). The Children’s Health Study is Southern California, not the San Joaquin Valley, but it is the canonical CHS pesticide-asthma result and is widely cited in California environmental health work.
These five studies, taken together, describe what California’s exposure data has been correctly used to find. The pattern is that sulfur-related and organophosphate exposures carry the clearest pediatric respiratory signal, fumigants are mixed, and the residential drift pathway is empirically real even when the outcome signal is uneven. This is more nuanced than the bumper-sticker version of the agricultural-pesticide-and-asthma story. It is also closer to what the data actually shows.
An in-house overlay attempt, and what it did not produce
We attempted a sub-county overlay of pesticide application intensity against modeled adult asthma prevalence from CDC PLACES in Fresno County. The attempt did not produce a publishable finding, and we are including the methods story honestly because the reasons for the null result are themselves informative for how CHNA practitioners should think about this kind of analysis.
The mechanics: we derived approximate centroids for each of Fresno’s 2,264 distinct PLSS sections from the Mount Diablo Meridian PLSS grid, snapped each section centroid to the nearest 2020 Census tract centroid, summed pesticide pounds to tracts, divided by tract land area to get an intensity measure, and compared modeled adult asthma prevalence (CDC PLACES 2024 release, data year 2023) between the top quintile and bottom quintile of pesticide intensity tracts.
The result: top-quintile-pesticide tracts had a mean modeled asthma prevalence of 9.95%, and bottom-quintile-pesticide tracts had a mean of 10.20%. The correlation between intensity and asthma across all 225 Fresno tracts was effectively zero (Pearson r = -0.018; Spearman = -0.235, weak negative). The high-pesticide tracts were rural (median 973 people per square mile) and the low-pesticide tracts were urban (6,109 per square mile). The CDC PLACES tract-level modeled estimates for Fresno have a standard deviation of only 0.65 percentage points across all 225 tracts, which is to say there is barely any spatial variation in the modeled surface to begin with.
Three things are wrong with this overlay. First, the PLSS sections do not nest into census tracts; we used approximate centroid snapping rather than rigorous area-weighted attribution because the platform does not currently carry CDPR’s PLSS section geometry. Second, the CDC PLACES tract-level adult asthma estimates are model-based small-area estimates driven by demographic predictors, so overlaying them against pesticide intensity is partially circular: both surfaces are demographically driven. Third, and most importantly, modeled adult asthma is not the right outcome to look at if we believe the pesticide-respiratory pathway is real. The CHAMACOS cohort work that found signal was pediatric, longitudinal, and used measured exposure with confounder control.
What a credible spatial overlay in Fresno County would require: California HCAI hospitalization and emergency department data geocoded to ZIP or census tract (these are measured events, not modeled prevalences); a real exposure surface using CDPR’s PLSS section geometry with inverse-distance or dispersion-weighted attribution within 1 to 4 kilometers of population; confounder control for age, race, poverty, urbanicity, and ideally occupational status; and restriction to populated tracts with meaningful population near treated sections. The Raanan and Gunier work used substantively this approach. The platform does not currently carry HCAI data or PLSS section geometry.
The honest conclusion is that California’s section-level PUR data is uniquely valuable as an exposure-context input for CHNA work. It is not, on its own, sufficient to demonstrate exposure-response relationships in a county-level CHNA. The exposure-response evidence comes from the published epidemiological literature, which has already done this work correctly in a different California cohort.
What California CHNA practitioners can do with section-level PUR data right now
For FQHCs, county and district public health departments, critical access hospitals, and Medi-Cal managed care plans operating in California’s agricultural counties, the section-level PUR data supports several specific CHNA workflows that would not be possible with county-aggregate data.
A CHNA can describe the specific PLSS sections, application months, and active ingredients within the catchment served by each clinic site or hospital. This is exposure context for the population served, not a causal claim. It supports targeted respiratory screening placement, promotores outreach timing aligned to fumigation seasons, and service-area justification for HRSA Section 330 grant defensibility.
A CHNA for a Critical Access Hospital like Coalinga Regional Medical Center (Fresno County’s only CAH, sitting in the heavily farmed western valley) can describe the high-intensity sections within the hospital’s primary service area and align implementation strategies (respiratory screening protocols, fumigation-season emergency department staffing, community notification tie-ins with CDPR’s SprayDays California pesticide-application notification system) to the specific spatial pattern.
A Medi-Cal managed care plan operating in Fresno (CalViva Health, Anthem Blue Cross), Kern (Kern Family Health Care, Anthem), or Tulare (Anthem, Health Net) can stratify asthma and COPD members by residential proximity to high-intensity sections and target environmental-trigger management interventions accordingly. This is population-health stratification using exposure context, not a causal claim about individual plan members.
A 330(g) migrant and seasonal worker health center can document the specific seasonal application patterns near worker housing locations to support service-area needs assessments and targeted outreach. Clinica Sierra Vista, United Health Centers of the San Joaquin Valley, and Family HealthCare Network are the major Central Valley FQHC systems with documented farmworker-serving missions.
In each of these workflows, the section-level data adds analytical capability that is operationally impossible in any other state. A CHNA practitioner in Iowa, Mississippi, or Indiana cannot describe exposure context at the square-mile level because the data does not exist. California CHNA practitioners can. This is the structural fact that should shape methodology choices for California CHNAs in the 2026–2028 cycle.
SprayDays California and the broader regulatory context
California’s regulatory environment around pesticide notification and use has shifted substantially in the past two years. SprayDays California, the statewide pesticide-application notification system, went live on March 24, 2025, providing 48-hour notice for soil fumigant applications and 24-hour notice for other restricted-material applications. The first annual update was published March 10, 2026. CDPR’s Regulation 22-005 for 1,3-dichloropropene took effect January 1, 2024, with buffer zone requirements, exclusion periods, totally impermeable film tarp requirements, and annual reporting obligations. DPR 24-001, also addressing 1,3-D health risk mitigation, takes effect January 1, 2026.
For CHNA implementation strategies in California’s agricultural counties, the regulatory shifts are part of the implementation context. A 2026 implementation strategy that names SprayDays California as part of a community notification approach, that references the 1,3-D buffer zone framework when describing population-near-fumigation interventions, and that ties promotores outreach to the published regulatory restricted-material list is using the regulatory environment as part of the strategy. A 2026 implementation strategy that is silent on this regulatory layer is leaving operational tools on the table.
How Banana Analytics fits this workflow
The Banana Analytics platform ingests California PUR at the section grain for the years currently available (2020–2022, with 2023 ingestion pending an identified pipeline defect), alongside CDC PLACES, CDC WONDER, HRSA NPPES, Census ACS, EPA AQS, EPA TRI, FEMA NRI, First Street Foundation, CMS Geographic Variation, and other federal and commercial sources. For California CHNA work, the platform surfaces section-level pesticide use as part of the Environmental Risk dimension, integrated with the broader compound signal scoring across all 3,222 US counties and roughly 74,000 census tracts.
What the platform automates for California CHNA workflows: pulling the section-level application records for a defined community (county selection, drive-time isochrone, attribution file, or state boundary), computing intensity rankings within the community, identifying the top compounds and their toxicological context, integrating the exposure context with CDC PLACES respiratory measures and HRSA provider supply data, and generating a methodology section with cited sources and vintages. The analytical work that would otherwise take an analyst several days to assemble manually becomes a defined workflow with citation discipline that holds up under audit.
The HCAI hospitalization and emergency department data, CHIS childhood asthma estimates, and PLSS section geometry that would support a rigorous in-house spatial overlay analysis are not currently on the platform. We have flagged these as the highest-value data additions for California-specific CHNA work and the most direct path to enabling in-house exposure-response analysis at the sub-county level.
Tier details and pricing are published openly on the pricing page. We have a structural commitment to access for organizations serving underserved populations: if your organization serves underserved communities and a paid license is genuinely out of reach, we provide Professional or Studio access at no cost. That commitment is not a promotional offer. It is structural, like the PBC incorporation and the 1% for the Planet pledge.
A short reference list
The peer-reviewed studies cited above:
- Raanan R, Gunier RB, Balmes JR, et al. (2017). Elemental Sulfur Use and Associations with Pediatric Lung Function and Respiratory Symptoms in an Agricultural Community (California, USA). Environ Health Perspect 125(8):087007. DOI: 10.1289/EHP528
- Raanan R, Harley KG, Balmes JR, et al. (2015). Early-life Exposure to Organophosphate Pesticides and Pediatric Respiratory Symptoms in the CHAMACOS Cohort. Environ Health Perspect 123(2):179–185. DOI: 10.1289/ehp.1408235
- Gunier RB, Raanan R, Castorina R, et al. (2018). Residential Proximity to Agricultural Fumigant Use and Respiratory Health in 7-year-old Children. Environ Res 164:93–99. DOI: 10.1016/j.envres.2018.02.022
- Salam MT, Li YF, Langholz B, Gilliland FD (2004). Early-life Environmental Risk Factors for Asthma: Findings from the Children’s Health Study. Environ Health Perspect 112(6):760–765. DOI: 10.1289/ehp.6662
- Madrigal JM, Gunier RB, Jones RR, et al. (2024). Residential Proximity to Agricultural Herbicide and Fungicide Applications and Dust Levels in Homes of California Children. Environ Int 192:109024. DOI: 10.1016/j.envint.2024.109024
The data and regulatory sources cited:
- California Department of Pesticide Regulation, Pesticide Use Reporting (PUR) data via the CalPIP portal
- SprayDays California pesticide application notification system
- CDPR Regulation 22-005 for 1,3-Dichloropropene
- USGS Pesticide National Synthesis Project, the federal alternative used for non-California analysis