Spring planting is forgiving. If you get your lettuce in ten days late, the days are still getting longer and warmer, and the plant catches up. Fall is the opposite. Every day you delay costs you more than a day of growth, because the plant is racing a clock that speeds up as it runs. Sow your fall broccoli two weeks late and you will not harvest broccoli — you will harvest disappointment, in the form of thumb-sized heads under a tarp in November.

The fix is not complicated, but it is arithmetic, and it has to be done before the season, not during it. You count backward from your average first frost date, apply a correction factor for shortening days, and work out an actual sow date for every crop you want in the ground. Then you follow the calendar even when it feels absurd to be seeding brassicas in the middle of a July heat wave.

Find Your Real First Frost Date

Start with the published average first frost date for your area — usually available from your regional extension service or any frost date lookup keyed to your ZIP code. That date is a 50% probability: half the years frost arrives earlier, half later. For planning purposes, that is a coin flip, and coin flips are a bad basis for a harvest.

Two adjustments matter. First, shift your planning date 10 to 14 days earlier than the published average if you want roughly 80–90% confidence. Second, correct for your actual site. A garden at the bottom of a slope, in a low spot where cold air pools, frosts 5 to 10 days earlier than the regional average. A garden on a south-facing slope, near a masonry wall, or in an urban heat island frosts later — sometimes two weeks later.

The only way to know your site is to record it. Keep a running log of the first light frost (patchy, 32°F), the first killing frost (28°F or below), and what actually died on each date. After three seasons you will have a personal frost date far more useful than anything published, and you will stop trusting the map. This is exactly the kind of multi-year observation that gets lost in a notebook and survives in an app — CropsBook lets you log dated field observations against specific beds so the pattern is still there when you need it next August.

Your first frost date is not a wall, it is a filter. Tender crops stop at that line. Hardy crops walk right through it and keep going for another six weeks.

The Fall Factor: Why Days to Maturity Lies

Seed packets list days to maturity based on spring and summer conditions — long days, warming soil, plants growing at full tilt. In fall, daylength is shrinking by two to three minutes a day, light intensity is dropping, and soil temperature is falling. The same crop takes meaningfully longer.

The standard correction is the fall factor: add roughly 14 days to the packet's days to maturity for anything maturing in September or October. Some growers scale it — add 10 days for early-September maturity, 14 for late September, and 20 or more for anything targeting late October, when growth nearly stalls.

Then add the other buffers people forget:

  • Germination lag — direct-sown seed in hot late-summer soil may take 4–7 extra days, or fail entirely above 80°F for lettuce and spinach. Count germination as its own block of days, not part of DTM.
  • Transplant shock — add 7–10 days if you are setting out transplants, because packet DTM for transplanted crops is usually counted from transplant date, but a stressed August transplant sulks for a week before it moves.
  • Harvest window — you rarely want one head of cabbage on one day. Add 14 days so you have a real picking window rather than a single date.

Put together, the formula is: first frost date − (days to maturity + fall factor + germination lag + harvest window) = sow date. For a 65-day broccoli with an October 10 frost date, that is 65 + 14 + 7 (transplant) + 14 = 100 days, which puts your transplant-to-field date around July 2 and your seed-starting date around June 1. That is why fall gardening feels wrong — the work happens in the hottest part of the year.

Sort Crops by Frost Tolerance, Not by Preference

The countdown above only applies to crops that must finish before frost. Many fall crops do not. Sorting your list into tolerance tiers changes the math completely.

  • Tender (dies at 32°F) — beans, cucumbers, summer squash, corn, basil. These must be fully mature before your adjusted frost date. No exceptions, no rescue.
  • Semi-hardy (survives to about 28°F) — lettuce, chard, beets, carrots, cauliflower, celery, potatoes (tops). You get roughly two extra weeks past first frost. Plan maturity for the frost date and treat the extra weeks as insurance.
  • Hardy (survives 24–28°F repeatedly) — kale, collards, cabbage, broccoli, leeks, turnips, radishes, spinach, mache, parsnips. Plan for these to reach near-maturity by frost, then hold in the field for four to eight more weeks.
  • Very hardy (survives into the teens with cover) — mature spinach, mache, claytonia, some kale varieties, overwintered garlic and shallots. These are less about a harvest date and more about a standing crop you cut from all winter.

The critical insight for hardy crops: they need to reach about 75% of their mature size before the "Persephone period" — the stretch when daylength drops below 10 hours and plant growth essentially stops. Depending on latitude, that begins somewhere between early November and early December. After that point, nothing you plant grows; it only holds. So the real deadline for hardy greens is not frost at all, it is the 10-hour-day line.

Ready to put this into practice? Download CropsBook on the App Store — it’s free and works offline.

Build the Actual Calendar

Take a page and write your adjusted frost date at the top. Then, for each crop, write the total day count and the resulting sow date. A typical zone 6 list, planning against an October 10 frost:

  • Brussels sprouts (100 days) — seed indoors mid-May, transplant late June. The longest lead time in the fall garden and the one people miss most often.
  • Cabbage, broccoli, cauliflower (60–75 days) — seed late May to mid-June, transplant early to mid-July.
  • Leeks (100–120 days) — seed in February or March; these are effectively a spring-sown fall crop.
  • Carrots and beets (55–70 days) — direct sow mid-July to early August. Carrots sweeten dramatically after frost, so err late rather than early.
  • Kale, collards, chard (50–60 days) — sow mid-July for full-size plants, or early August for a smaller standing crop.
  • Bush beans, summer squash (50–55 days) — direct sow no later than mid-July. Tender, so the frost date is a hard stop.
  • Spinach (40–45 days) — sow late August through mid-September; germination is unreliable above 70°F soil.
  • Lettuce (45–55 days) — sow every two weeks from early August to mid-September for continuous cutting.
  • Radishes, arugula, mustards (25–35 days) — sow through late September; these are your gap fillers wherever summer crops come out.
  • Garlic — plant 2–4 weeks after first frost, for harvest the following July.

Notice that half these dates fall in June and July. The fall garden is not something you decide on in September. It is something you commit to in early summer, when the spring beds are still full and finding space feels impossible.

If you are not seeding fall brassicas while you are still harvesting spring peas, you are already behind.

Solve the Space Problem Before It Arrives

The genuine obstacle to fall planting is that the beds are occupied. Your July transplant date collides with tomatoes that will produce until October. Three practical answers:

Nursery beds. Grow fall transplants in a small dedicated bed or in trays at high density, then move them into place the day a summer crop comes out. A 4×4 nursery bed can hold 200 brassica transplants for six weeks. This decouples your sow date from your bed availability, which is the single highest-leverage change you can make.

Ruthless removal. The cucumber vines are still alive in late July, but they are producing four bitter fruits a week and hosting downy mildew. Pull them. A bed of fall carrots planted July 20 is worth far more than three more weeks of a failing cucumber crop. Track yield per bed and this decision stops being emotional — it becomes obvious from the numbers.

Interplanting. Set fall brassica transplants between determinate tomatoes or under the trellis of pole beans in mid-July. The summer crop provides shade during establishment; by the time the brassicas need full light, the summer crop is done and pulled.

All three depend on knowing what is planted where and when it went in. If you are running more than three or four beds, memory fails around week six. Whether you use a paper map or something like CropsBook to keep bed-by-bed planting records, the point is that the fall plan requires knowing your summer plan in detail. The same discipline applies across small-farm operations generally — growers running livestock alongside produce track breeding and feed schedules in Barnsbook, and beekeepers time their fall feeding and mite treatments off similar backward-counted calendars in HiveBook. Different animals, same arithmetic: work backward from the hard deadline nature sets.

Get Seed to Germinate in Hot Soil

Your July and August sowings go into soil that may be 85°F or hotter. Lettuce enters thermal dormancy above 80°F. Spinach germination collapses above 75°F. Carrots germinate fine in warm soil but dry out and die in the top half inch within hours.

  • Water the furrow, not the bed — soak the open furrow, sow into wet soil, cover with dry soil. This puts moisture where the seed is and leaves a dry crust that does not cake.
  • Shade the row — a board laid directly on a carrot row, checked daily and removed at first emergence, drops soil temperature 10–15°F and holds moisture. Shade cloth at 30–40% over the bed does the same for larger plantings.
  • Prime lettuce and spinach seed — soak 4–6 hours, drain, then hold at 40°F in a damp paper towel for 24–48 hours before sowing. This breaks thermal dormancy and gets you germination in soil that would otherwise refuse.
  • Sow deeper than spring — a half inch instead of a quarter for small seed, reaching cooler, wetter soil. Increase the seeding rate 25–50% to compensate for losses.
  • Start in trays where you can — brassicas, lettuce, and chard all transplant well. Germinating them in a shaded, controlled spot beats fighting an August bed.

Extend the End of the Season

Once frost arrives, your goal shifts from growing to holding. A single layer of medium-weight row cover over hoops buys 4–6°F of protection and, more importantly, blocks radiational cooling on clear still nights — which is when frost damage actually happens. A low tunnel inside an unheated high tunnel gives you roughly two USDA zones of shift and lets zone 6 growers cut spinach in January.

Timing matters more than material. Cover before the first frost, not after — a plant that has been hardened by cool nights in the 40s handles 28°F far better than one shocked straight from 60°F nights. And vent on sunny days: a closed low tunnel can hit 90°F in October sun and cook lettuce you spent three months growing.

Row cover does not create heat. It traps what the soil already has. Wet, dark, uncovered soil stores more daytime warmth than dry mulched soil, so water your beds before a hard frost night.

For root crops, the simplest storage is no storage: leave carrots, parsnips, and leeks in the ground under 12 inches of straw and harvest through winter as long as you can get a fork in. Parsnips and carrots both convert starch to sugar after repeated frost, so the January harvest genuinely tastes better than the October one.

Record This Year So Next Year Is Easier

The first fall calendar you build will be wrong in specific, learnable ways. Your brassicas will be two weeks behind, or your spinach will bolt, or the frost will come October 1 instead of October 10. None of that matters if you write it down.

Log four things per planting: sow date, actual germination or transplant date, first harvest date, and what the weather did in between. The gap between predicted and actual maturity is your personal fall factor — and it will be more accurate than the generic 14 days. After two seasons you will be adjusting by crop and by bed rather than by rule of thumb.

The fall garden rewards planning more than any other part of the season, because there is no recovery from being late. Spring gives you a second chance; fall gives you frost. Build the calendar in June, put the dates somewhere you will actually see them, and plant on the date even when the thermometer says it is madness. The reward is the part of the season nobody else gets — carrots sweeter than anything you grew in July, spinach cut in a January thaw, and a garden still producing when the neighbors' beds have been bare since Labor Day.