You want clover in the pasture, the orchard rows, or the thin patch out back, without renting a no-till drill or tearing up established sod. Frost seeding solves exactly that: broadcast small clover seed onto frozen ground in late winter, and the ground plants it for you.
The mechanism is simple. When soil freezes at night and thaws by day, the surface lifts into a honeycombed crust of ice needles, then collapses. Seed sitting on that surface drops into the gaps. Penn State Extension describes this honeycombing of the soil surface as the whole reason the practice works, and recommends broadcasting early in the morning while the ground is still hard, with a thaw expected later that day.
USDA NRCS calls it one of the most cost effective seeding methods available, and notes that because it works on slopes it avoids the erosion tilling would cause. It is honest about the limits too: the probability of success drops in dry springs and on sandy soils, where there is neither enough soil movement to bury seed nor enough moisture to carry a seedling.
Below: the window for your region, seeding rates converted for small plots, the species extension tells you not to frost seed, the four reasons stands fail, and what a good stand returns in nitrogen.
Key Takeaway
Frost seeding is a clover practice, not a general seeding method. On grazed-down ground with loam or clay soil, broadcasting 4 to 8 pounds of red clover per acre in February costs an hour and a bag of seed. On thick thatch, sandy ground, or after the heaving cycles have finished, it is money left on the surface.
The calendar date is a proxy. The real signal is nightly freezing, daytime thawing, and enough weeks before green-up for a seedling to establish. Extension services across the eastern half of the country converge on late winter, but the window shifts by several weeks between the Upper South and the Upper Midwest.
| Region | Recommended window | Source |
| Kentucky and transition zone | February 10 to March 1, mid-February preferred | UK Master Grazer |
| Missouri and Upper South | January or February | University of Missouri Extension |
| Ohio and Lower Great Lakes | February for most of the state | Ohio State Extension, 2026 |
| Iowa and Upper Midwest | February and early March, ground still frozen | Iowa State Extension |
| Mid-Atlantic and Northeast | February and March, seeded in the morning | Penn State Extension |
Sources: UK Master Grazer, University of Missouri Extension, Ohio State C.O.R.N. Newsletter, Iowa State Extension, Penn State Extension.
Notice what is missing: a soil temperature threshold. No extension service in this set publishes one, because the useful measurement is visual. If frost heave is lifting the soil in the morning and the ground is soft by mid-afternoon, you are in the window. If it has gone soft and stayed soft, you have missed it, and an autumn sowing is the better route. Our guide to winter cover crops covers the species for that.
NRCS advises against seeding onto heavy snow, and Ohio State goes further: avoid any area where bare dirt was not visible before the snow fell. A dusting that melts by noon is workable. Six inches (15 cm) of pack is a wasted bag.
Medium red clover is the species to start with. NRCS singles it out as the easiest legume to frost seed, for its seedling vigor and cold tolerance. White and ladino clover work too, at lower rates, and suit a low persistent ground layer rather than a tall stand you plan to cut. Broadcast rates run higher than drilled rates because some seed never reaches soil.
| Species and context | Rate (lb per acre) | Source |
| Red clover, pasture | 4 to 6 (4.5 to 6.7 kg/ha) | USDA NRCS, 2022 |
| Red clover, renovation | 4 to 8 (4.5 to 9.0 kg/ha) | Penn State Extension |
| Ladino white clover | 2 to 3 (2.2 to 3.4 kg/ha) | NRCS and Penn State |
| Red plus ladino mix | 3 to 4 red plus 1 to 2 ladino | USDA NRCS, 2022 |
| Kentucky renovation mix | 6 to 8 red plus 1 to 2 ladino | Kentucky AGR-271 |
| Alsike clover | 1 to 4 (1.1 to 4.5 kg/ha) | Iowa State Extension |
| Birdsfoot trefoil | 4 to 6 (4.5 to 6.7 kg/ha) | Penn State Extension |
Sources: USDA NRCS Frost Seeding, Penn State Extension, University of Kentucky AGR-271, Iowa State Extension.
Most people here are not measuring in acres. An acre is 43,560 square feet, so 4 pounds per acre is roughly 1.5 ounces per 1,000 square feet (about 43 g per 93 sq m), and 8 pounds per acre about 3 ounces. That is arithmetic, not an extension recommendation. Use a kitchen scale, because at these rates a doubling error is easy and expensive.
Do Not Frost Seed These
The University of Kentucky states plainly that frost seeding is not recommended for grasses or alfalfa. Alfalfa seed is too large and needs deeper, controlled placement. NRCS does list ryegrass at 8 to 10 lb per acre and orchardgrass at 3 to 4, while warning that grass success is consistently lower. Timothy, tall fescue and Kentucky bluegrass have no extension backing for this method at all.
Test and fix soil pH the previous summer
Clover wants pH 6.1 to 6.7. Master Grazer advises liming low-productivity ground six months ahead, because lime is slow. Below pH 6.0, Maine Extension notes lime is not optional if you want nitrogen fixation. Start with a soil test, then raise pH with lime.
Open the ground in autumn
Graze hard or mow close in fall so residue is thin and soil visible. NRCS is explicit that better stands come from thin sods and bunch grasses than thick sod-forming turf. Most people skip this step, and it decides the outcome.
Buy inoculated seed, or inoculate it yourself
Red clover needs Rhizobium leguminosarum biovar trifolii to nodulate. Kentucky advises pre-inoculated seed or treating it before sowing, and warns against "common" or variety-not-stated seed. No published data shows coated seed lets you cut the rate, so do not.
Broadcast on a frozen morning
A hand-crank chest spreader covers a quarter acre; a cyclone spreader on an ATV handles more. Calibrate on a tarp first, because at 4 pounds per acre a spreader set two notches wide empties the bag in a third of the area. Walk two passes at right angles.
Leave it alone until seedlings can take traffic
Kentucky and Ohio State both warn against grazing frost-seeded ground until clover is tall enough to survive hoof traffic. Skip broadcast herbicide entirely, since most broadleaf products kill clover seedlings. Handle weeds the autumn before, or spot-spray later.
Why This Works: Fertility Bought With Timing, Not Inputs
Frost seeding puts a legume where a fertilizer bill used to be, and it does it in February, when nothing else on the place needs attention. A good red clover stand at 30 to 40 percent of the sward fixes 100 to 200 pounds of nitrogen per acre per year. You bought that with a bag of seed and an hour of walking, in the one month when labour costs you nothing.
Get Our Free Companion Planting Chart
Join 10,000+ gardeners getting weekly tips on what to plant together, soil health, and permaculture techniques.
Send Me the ChartExtension services name four causes, and they do not weight them equally. Poor seed-to-soil contact appears in every publication, which tells you where to spend attention. Thick thatch, undisturbed sod-forming grass, or residue left standing over winter all stop seed reaching mineral soil. Penn State suggests letting livestock tramp seed in where thatch is unavoidable.
Second is soil texture. Penn State says the method works on loamy and clay soils that hold water, and is not suited to sandy or shaley ground that dries fast. Ohio State adds the mechanical reason: dry sandy soils move less during freeze-thaw, so less burial happens.
Third is timing, the failure that feels like bad luck and is not. Seed broadcast after the heaving cycles end simply sits there. Fourth is what follows: grazing too early, or a broadcast herbicide pass that kills the seedlings you paid for. Worth saying plainly, no extension service in this evidence set publishes a numeric success rate for frost seeding, or a failure rate by cause. Treat it as a good bet with real variance, and plan to reseed thin patches next year.
Key Takeaway
If you change one thing, graze or mow short in autumn. Seed-to-soil contact is the failure named in every publication, and the only one entirely within your control. Weather, soil texture and spring rainfall are not.
Nitrogen, mostly, and more than people expect. Published figures vary widely because stand density and season length vary widely, so here is the range rather than a single number.
| Legume and context | Nitrogen fixed (lb per acre per year) | Source |
| Red clover at 30-40% of a mix | 100 to 200 | Mississippi State Extension, 2018 |
| White clover at 30-40% of a mix | 100 to 150 | Mississippi State Extension, 2018 |
| Mammoth red clover, forage stand | 70 to 110 | Maine Cooperative Extension, 2017 |
| Red clover, general range | 50 to 250 | Georgia Forages, 2013 |
| White clover, general range | 74 to 252 | Georgia Forages, 2013 |
| Transferred to neighbouring grass | 2 to 38 | Missouri Forage and Grassland Council, 2024 |
Sources: Mississippi State Extension, University of Maine Extension, Georgia Forages, Missouri Forage and Grassland Council.
Read that last row carefully, because it is the one that gets oversold. Only a fraction of what clover fixes reaches the grass beside it, often no more than 10 to 20 percent while the stand is alive. The rest stays in roots and tops until they die back. That still matters, and it is the same mechanism at work in every nitrogen fixing plant you grow, but it is a slow-release account rather than a transfer.
The numbers get concrete at termination. University of Minnesota Extension trials found red clover without a companion crop produced roughly 1,320 pounds of fall herbage and 1,952 pounds of root dry matter per acre, carrying about 40 pounds of nitrogen in herbage and 50 in roots, ahead of corn yielding around 163 bushels per acre. One caution: once clover exceeds about 25 percent of a sward, stop applying nitrogen fertilizer, which suppresses nodulation. The same logic applies in an orchard understory, where clover slots into the ground cover layer of a food forest.
Frost seeding is broadcasting seed onto the soil surface in late winter, while the ground is still frozen, and relying on repeated freezing and thawing to pull it into shallow contact with soil. USDA NRCS and Penn State Extension both define it this way. No tillage or covering is involved. It works because freeze-thaw honeycombs the surface, opening gaps that small seed falls into before the surface collapses. It is used to renovate thin pastures, add clover to existing sod, and establish cover crops into standing wheat or barley.
Roughly February 10 to March 1 in Kentucky and the transition zone, mid-February preferred. January or February in Missouri, February in most of Ohio, and February into early March across Iowa, the Northeast and the Mid-Atlantic. The date matters less than the condition: ground that freezes at night, thaws by day, and has several such cycles ahead. Sow in the morning while the surface is hard, so you avoid rutting soft ground and still get the heaving you need.
Extension services say no. The University of Kentucky explicitly excludes alfalfa, alongside grasses. Alfalfa seed is larger than clover seed and needs deeper, more controlled placement than surface broadcast and frost heaving can give it. There is also the autotoxicity problem when seeding alfalfa into an existing alfalfa stand. If alfalfa is the goal, drill it at the correct depth in spring or late summer, and keep frost seeding for red, white and alsike clover.
Crimson clover is small-seeded and cool-tolerant, so it fits the profile, but it is an annual and far less commonly recommended for this than medium red clover. NRCS names medium red clover as the easiest legume to frost seed, and the published rate tables cover red, ladino, alsike and sweet clover rather than crimson. If crimson is what you want, an autumn sowing is better supported. Our guide to crimson clover as a cover crop covers that timing.
Less reliably than clover, and several extension services recommend against it. Iowa State attributes the difference to seed size, since grass seed needs more soil movement to bury properly. NRCS publishes rates for forage ryegrass at 8 to 10 pounds per acre and orchardgrass at 3 to 4, while noting success is lower than for legumes. Tall fescue, timothy and Kentucky bluegrass have no extension support at all, so drilling in spring or late summer is safer. Building soil first, as our soil health guide lays out, improves the odds either way.
Ready to Grow Smarter?
Get our free 20-page beginner's guide to backyard food forests. Two printable worksheets, a year-by-month food-forest calendar, and a curated reading path.
Read the Free Guide