Total US farm pesticide use is up about 21% since GMOs arrived, not "dramatically" — the real story is a 9x shift toward glyphosate as insecticide use fell

Summary

The premise needs correcting before the number: total US farm pesticide use has NOT "dramatically" increased since GMOs arrived in 1996 — it is up about 21% at the national level (from a 1992-95 pre-GMO baseline average of ~475 million kg/yr of active ingredient to a 2017-18 average of ~572 million kg/yr, per USGS county-level estimates). What actually changed dramatically is the composition: glyphosate use rose roughly 9-fold (+838%) as glyphosate-tolerant ("Roundup Ready") soy, corn and cotton spread, while several major insecticides fell 40-99% as Bt corn and cotton reduced the need to spray for corn borer, bollworm and rootworm. The modest aggregate increase is the net of a large herbicide gain and a real insecticide decline — a substitution story, not a uniform escalation.

What the data shows

Queried directly from ag.pesticide_use_by_county (USGS NAWQA Pesticide National Synthesis Project, county-level agricultural pesticide-use estimates, active-ingredient kilograms, 1992-2019 observed coverage):

• Total US agricultural pesticide use: 1992-95 average ≈ 475 million kg/yr (≈1,047 million lb/yr) → 2017-18 average ≈ 572 million kg/yr (≈1,262 million lb/yr). That is a +21% increase over roughly two decades of GMO adoption — real, but not dramatic at the aggregate level.
• Glyphosate (the herbicide paired with Roundup Ready crops): 1994-95 average ≈ 13.4 million kg/yr → 2017-18 average ≈ 125.7 million kg/yr, a 9.4x increase (+838%). Glyphosate alone accounts for most of the growth in total herbicide mass.
• Representative insecticides displaced by Bt crops: chlorpyrifos -42%, carbaryl -83%, terbufos -80%, aldicarb -92%, methyl parathion -99% (comparing the same 1994-95 vs 2017-18 windows).

Data-quality note on the endpoint year: 2019 shows an apparent sharp drop in this table (total falls to ~341M kg) that is a reporting-coverage artifact, not a real decline — the source's own documentation says releases after the 2013-17 consolidated file are "preliminary" and cover fewer compounds. For that reason the "recent" comparison year used here is 2017-18 (the last fully finalized release), not 2019. The table's observed coverage window is 1992-2019; nothing more recent is loaded, so "today" in this analysis means the most recent reliable USGS estimate (2018), not the current year.

Why total mass understates the herbicide story

Published research on this question reaches the same conclusion this data does. Charles Benbrook's peer-reviewed analysis (Environmental Sciences Europe, 2012) found that herbicide-tolerant GMO crop technology added about 239 million kg (527 million lb) of herbicide use in the US between 1996 and 2011, while Bt insect-resistant crops reduced insecticide use by about 56 million kg (123 million lb) over the same period — a large herbicide increase partly offset by an insecticide decrease, netting to a smaller aggregate change. A separate USDA-NASS-sourced estimate cited in that literature found glyphosate use on soybeans in 2016 was about 13x higher than in 1995, matching the magnitude found independently here (9.4x nationally across all crops, 1994-95 to 2017-18).

Since 2016, glyphosate-and-dicamba-tolerant soy and cotton have also driven a rise in dicamba use (21.5 million kg in 2017-18 in this data, essentially new), adding a second herbicide-specific growth driver not present before GMOs.

Bottom line

If the question is "how much more pesticide mass is applied overall": about 21% more today than in the early-1990s pre-GMO baseline, per USGS estimates (1992-95 avg ~475M kg/yr vs 2017-18 avg ~572M kg/yr). If the question is about the herbicide GMO crops are specifically engineered to work with: glyphosate use is up roughly 9-fold. The "dramatic increase" framing in the question is only accurate for that one compound class; it overstates what happened to total pesticide use, because it omits the substantial, GMO-driven decline in insecticide use that the same technology produced on the same acres.

What This Report Does Not Answer

• 'Today' — most current available data: Table's observed coverage ends 2019, and 2019 itself is a preliminary/incomplete release showing an artifactual drop; used 2017-18 (last finalized data) as the most recent reliable point instead of true present-day 2026 figures, which are not loaded in this corpus.

Every query behind this report

8 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT type, count(*) c FROM ag.pesticide_use_by_county GROUP BY type ORDER BY c DESC

SELECT "year", SUM(epest_high_kg) AS total_high_kg, SUM(epest_low_kg) AS total_low_kg
FROM ag.pesticide_use_by_county
WHERE "year" BETWEEN 1992 AND 2019
GROUP BY "year"
ORDER BY "year"

SELECT compound, SUM(epest_high_kg) AS kg
FROM ag.pesticide_use_by_county
WHERE "year" IN (1994,1995)
GROUP BY compound
ORDER BY kg DESC
LIMIT 15

SELECT compound, SUM(epest_high_kg) AS kg
FROM ag.pesticide_use_by_county
WHERE "year" IN (2017,2018)
GROUP BY compound
ORDER BY kg DESC
LIMIT 15

SELECT compound, SUM(epest_high_kg) AS kg
FROM ag.pesticide_use_by_county
WHERE "year" = 2019
GROUP BY compound
ORDER BY kg DESC
LIMIT 15

SELECT "year", compound, SUM(epest_high_kg) AS kg
FROM ag.pesticide_use_by_county
WHERE compound IN ('METHYL PARATHION','CHLORPYRIFOS','CARBOFURAN','TERBUFOS','ALDICARB','CARBARYL')
AND "year" IN (1994,1995,2017,2018)
GROUP BY "year", compound
ORDER BY compound, "year"

SELECT "year", SUM(epest_high_kg) AS kg
FROM ag.pesticide_use_by_county
WHERE compound = 'GLYPHOSATE'
AND "year" IN (1994,1995,2017,2018)
GROUP BY "year"
ORDER BY "year"

SELECT "year", SUM(epest_high_kg) AS total_kg
FROM ag.pesticide_use_by_county
WHERE "year" BETWEEN 1992 AND 2018
GROUP BY "year"
ORDER BY "year"