Every packet of seed you buy started in someone’s garden. Saving your own is older than agriculture as an industry, and it is still one of the highest-leverage skills a vegetable grower can pick up. Done well, it cuts your seed bill to near zero, gives you stock adapted to your specific soil and climate, and protects you when a favorite variety vanishes from the catalog. Done sloppily, you end up with weak plants, cross-pollinated mystery hybrids, and moldy paper bags in the back of a drawer.

This guide walks through what to save, what to skip, how to harvest cleanly, and how to store seed so it actually germinates two or three years later.

Start With Crops That Make It Easy

Not every vegetable rewards a beginner equally. Some self-pollinate inside closed flowers and stay genetically stable with almost no effort. Others cross with anything in the neighborhood and need isolation distances measured in hundreds of feet. Start with the easy ones and earn your way up.

  • Tomatoes — self-pollinating, seeds inside the fruit, viable 4–6 years stored properly
  • Beans & peas — self-pollinating, dry on the plant, almost foolproof
  • Lettuce — self-pollinating, bolts and sets seed in one season
  • Peppers — mostly self-pollinating, some crossing risk between varieties planted close together
  • Eggplant — self-pollinating, easy harvest from overripe fruit

Avoid biennials (carrots, beets, onions, cabbage family) until you have a season of experience. They flower in year two, which means overwintering plants and managing far higher cross-pollination risk. Squash and cucumbers cross aggressively across varieties — technically easy but the results are unpredictable unless you hand-pollinate and tape flowers shut.

Open-Pollinated vs Hybrid: Why the Packet Matters

If the seed packet says F1 or hybrid, do not bother saving seed from it. Hybrid seed comes from a controlled cross between two parent lines. The first generation is uniform and vigorous. The second generation — what you would grow from saved seed — splits into a chaotic mix of grandparent traits. You might get one or two decent plants out of twenty.

Open-pollinated (OP) and heirloom varieties breed true. Save seed from a Brandywine tomato and you get Brandywine tomatoes next year. This is the foundation everything else rests on. Before you save anything, check the packet or catalog listing.

The cheapest insurance against a bad season is a tin of last year’s seed sitting in your closet. The cheapest insurance against losing a variety you love is saving it yourself, because catalogs drop varieties every year without warning.

Harvesting Wet-Seeded Crops the Right Way

Tomatoes, cucumbers, melons, and squash have seeds embedded in wet flesh. The seeds are coated in a gel that inhibits germination — useful inside a rotting fruit on the ground, less useful when you want to plant them. Ferment the gel off.

The process is simple. Scoop seeds and pulp into a jar. Add a splash of water. Leave at room temperature for 2–4 days, stirring once a day. A layer of mold will form on top. When viable seeds sink and pulp floats, pour off the junk, rinse the seeds in a sieve, and spread them in a single layer on a plate or coffee filter. Do not use paper towels — the seeds glue themselves down and tear. Dry until a seed snaps rather than bends, usually a week in a warm room.

Tracking which jar belongs to which variety matters more than you would think. Three jars of fermenting tomato seed look identical after two days. Label everything — variety, source plant, date. A simple notebook works, though tools like CropsBook make it easier to keep the planting record, the source plant ID, and the seed-save date linked together for next season’s planning.

Harvesting Dry-Seeded Crops

Beans, peas, lettuce, peppers, and grains dry on the plant. The discipline is patience — let them go further than feels right. A bean pod ready to save is brown, papery, and rattles when shaken. A lettuce head ready to save has bolted into a four-foot stalk topped with fluffy dandelion-like seed heads.

For beans and peas, pull whole plants when most pods are brown and hang them upside down somewhere dry for another week or two. Then thresh by hand or by stomping inside a pillowcase. Winnow by pouring the mix between two buckets in front of a fan — chaff blows out, seed falls.

For lettuce, the seed ripens unevenly. Bend the seed head over a paper bag every few days and tap. You will collect seed in waves over two weeks rather than all at once.

For peppers, just let a fruit go fully red, yellow, or whatever its mature color is, then cut it open and scrape seeds onto a plate to dry. No fermentation needed.

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

Drying and Storing Without Killing Viability

Heat and moisture kill seeds. Sun-drying on a windowsill seems intuitive and is one of the fastest ways to ruin a year’s work. Aim for temperatures under 95°F (35°C) during drying. A spare room with a fan moving air across the seeds beats any kind of direct heat source.

The standard rule is the sum of temperature in Fahrenheit plus relative humidity should stay under 100 for storage. A 70°F room at 25% humidity works. A 65°F basement at 50% humidity is borderline. Most home growers store in a closet or pantry and accept that viability drops faster than ideal — this is fine if you rotate stock annually.

  • Containers — glass jars with tight lids beat plastic bags for long storage. Paper envelopes inside a jar give you variety separation plus a moisture barrier.
  • Desiccant — a tablespoon of dry rice or a silica gel packet in each jar pulls residual moisture. Replace yearly.
  • Cool storage — the back of a closet beats the garage. A refrigerator extends viability for hard-to-replace varieties, but the seeds must be fully dry first or they crack from condensation.
  • Freezing — works for long-term storage of bone-dry seed, but every freeze-thaw cycle costs viability. Reserve for varieties you genuinely cannot replace.
Label twice. Variety on the envelope, variety on the jar. Six months from now, “tomato 2026” will not tell you whether it was the paste, the cherry, or the slicer you actually wanted.

Avoiding Cross-Pollination Disasters

Self-pollinating crops mostly take care of this for you. Tomatoes, beans, peas, lettuce, and peppers stay reasonably true even with multiple varieties planted close together. Some crossing happens in peppers and tomatoes, especially with bees around, but it is usually under 5%.

Insect-pollinated crops are the trap. Squash, cucumbers, melons, brassicas, and the carrot family will cross with anything related within a quarter mile. A zucchini next to a butternut produces seed that grows into something edible but weird. Brassicas are worse — broccoli, cabbage, kale, collards, kohlrabi, brussels sprouts, and several common weeds all interbreed.

Three ways to manage this. Grow one variety per species per year. Simplest. Isolate by distance, hundreds of feet for cucurbits, much more for brassicas. Hand-pollinate and bag flowers with mesh or tape the night before they open, then transfer pollen yourself in the morning. Tedious but works in tight spaces.

Beekeepers face the flip side of this same equation — if you run hives, the foraging radius of your bees affects whose seed crops cross with whose. Folks managing apiaries alongside vegetable plots often coordinate with neighbors, and dedicated tools like HiveBook help track hive locations and forage maps the same way CropsBook tracks planting blocks.

Selecting Which Plants to Save From

This is the part most home seed savers skip, and it is the part that actually improves your garden over time. Do not save seed from the first plant that produces. Do not save seed from the runt. Walk the patch with intent.

For each variety, pick three to five plants that show the traits you want more of — disease resistance, late bolting, heavy production, good flavor, the right ripening window for your climate. Tag them with flagging tape when you spot them. Save seed only from those plants.

Over five or six seasons of selecting this way, your saved seed becomes meaningfully better adapted to your soil, your rainfall, and your pest pressure than anything you can buy. This is how every regional landrace variety came to exist — gardeners selecting for what worked in their specific spot, year after year.

Avoid selecting from a single plant. Tomatoes can tolerate this, but most crops lose vigor from narrow genetics within a few generations. Five plants minimum, fifteen to twenty better for brassicas, corn, and other crops that need population-level diversity.

Building a Seed Library and Rotation System

Saved seed is only useful if you can find it and trust it. Build a simple system in year one and you will never lose stock to forgotten jars.

  • Test germination — before planting season, sprout ten seeds on a damp paper towel in a plastic bag. Count what comes up after a week. Anything under 70% germination means seed thicker or replace the stock.
  • Track viability windows — onions and parsnips drop off after one year. Most beans, peas, brassicas, and tomatoes hold 4–6 years. Lettuce, peppers, and carrots fall in the 2–3 year range.
  • Save more than you think you need — a pinch of lettuce seed plants a long row. A jar of bean seed is two seasons of insurance plus enough to share.
  • Date every envelope — not just “saved 2026” but month and crop stage. Helps later when troubleshooting low germination.

The administrative side — what you saved, when, from which planting, with what germination rate — is where this whole practice either becomes durable or collapses into a junk drawer. Keep it light. A notebook works. A crop log app works. The same record-keeping discipline applies to anyone managing biological inventory — growers using Barnsbook to track livestock lineages run on the same principle as a gardener tracking seed lines. Document the parent, document the date, document what you observed.

Seed saving is not really about seeds. It is about closing the loop on your garden so that next year’s success is something you built, not something you bought.

Start with three crops this season — pick a tomato, a bean, and a lettuce you actually liked eating. Save more seed than you need, label everything, store it dry and cool, and germination-test it next March. That single cycle teaches you more than any guide. Five seasons in, you will have a small library of varieties tuned to your dirt, your weather, and your taste — and a seed budget that rounds to zero.