Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Online Course Icon

Seeds and seed saving

Seed choices

It’s incredible to reflect on how much knowledge and growth power is contained in seeds. This lesson explains how you can unlock this potential, both by finding better seeds to buy and by saving seeds yourself. Homesaved seeds have qualities adapted to your soil and conditions, attributes that you can’t find in other seeds.

Sourcing seeds

As with many things in our fast-changing world, this is less straightforward than it used to be. Since the first Covid lockdowns, seed companies have been meeting unprecedented demand, both from new gardeners and from an expansion of vegetable growing at home. This has even caused many of them to ration their opening hours and restrict purchases.

In the UK, the issue is also complicated by Brexit, which is causing a reduction of the imports of seeds. The situation is too fluid for me to recommend suppliers, but I shall mention two in the UK: Real Seeds and the Seed Co-op.

The latter are based in an 8 hectare / 20 acre greenhouse in Lincolnshire, where they grow as many seeds as they can. They also import Bingenheimer seeds from Germany and, if you are in Europe, this is a company that I recommend, along with Sativa Seeds.

The ‘Seeds and Varieties’ page of this website, under the ‘Advice’ tab, has more advice on suppliers

All of these plants were grown from seed of the same packet – Palla Rossa chicory
Choosing seed – this is celeriac of two varieties, Giant Prague on the left and Ibis on the right; the former made a lot of leaves and smaller bulbs

Choosing suitable seeds

You need to read the small print on seed packets to discover things like harvest time, above all. Broccoli is a good example of this, because the word broccoli covers a multitude of possible results – harvests may occur any time from three to ten months after sowing. Some types, like the later varieties of purple sprouting broccoli, will grow large in autumn and then survive through winter before making their heads. Others are bred to flower before winter, from the same sowing date.

Sometimes there are surprises, and you are not to blame: seed is rarely uniform, except for F1 hybrids (see below).

Age of seeds

These photos show how much difference it makes, whether seed is old or young. It’s not only whether germination happens, but how fast it happens, and whether growth is stronger or weaker.

At least with homesaved seed, you know its age.

Homesaved Grenoble Red lettuce seed of different ages – one, two and three years old, with the youngest at the top
The effect of seed age on germination – one-year-old seed is behind and three-year-old seed is in front

I have had many experiences of poor growth from newly bought seed. The problem is that the information given is not about seed age, but when it was packeted!

Furthermore, what a seed company calls germination is not necessarily the same as an acceptable level of vigour in plant growth.

All of which is extremely poor, because we then waste time, space and resources trying to get old seeds growing, and potentially do not harvest a crop because of losing the early growth time. At least with propagation under cover, you are not suffering big empty spaces in beds, compared to the small spaces of seed and module trays.

From two different packets of Boltardy beetroot, sown 13 days previously – Seed Co-op on the left and Kings on the right, both bought a month earlier
This photo was taken in March 2018 – the Kings seed packet suggests young seed, but see the quote from Kings below*
Almost no germination of one of four batches of brassica seeds, all sown at the same time

* The man from Kings Seeds wrote to me on 21st March about what happened: ‘. . . germination was 85% at the time of packing the seeds, but a month later it suddenly fell to 72%, below acceptable standards’.

The legal minimum is 70% but I feel this means little, since it’s assessed in perfect laboratory conditions. 70% in a laboratory could be 20% in a garden. I would just add that I sow a lot of Kings seeds, and they are mostly good!

Saving seeds

When you save your own seeds, the difference is striking. They are much fresher and just want to grow, as long as you observe the basic tenets, such as plant numbers for cross-pollination. First off, however, I want to explain hybrid breeding.

F1 hybrids

Vegetable breeding in the last three or four decades has concentrated on hybrid varieties, many of which grow excellent vegetables, with high yields and good flavour. Unfortunately, from a seed saver’s perspective, they don’t breed true. For example, do not save seed of hybrid tomato varieties such as Sungold and Rosada F1. Plants I grew from ten seeds of Sungold grew tomatoes of varied shape and colour, and none were sweet or especially flavoursome.

  • F1 seeds are the result of firstly inbreeding two generations of plants to achieve desirable traits, then cross-pollinating the two lines, sometimes in a forced way.
  • The result is only one generation of a desired outcome and, if you sow seeds from those plants, your new plants have a random mix of less desirable traits.

Seed packets state ‘F1’ if they are hybrids. The process does not involve genetic modification, nor is it a natural method. The results are predictable, consistent and profitable, the reason for seed producers favouring work on F1 hybrids, rather than on open-pollinated strains (see Lesson 5 for a negative consequence of this).

New plants from sideshoots

For certain vegetables, another option for raising plants is vegetative reproduction, as with slips of sweet potato, the stem plus root of perennial onions, and sideshoots of tomato plants. The latter involves a fair amount of time, to keep them alive and healthy all winter. I do this with Rosada tomatoes because they are an F1 variety, so the seeds you save won’t grow true, and Rosada seeds are no longer available to buy.

November 16th – sideshoots of Rosada tomato plants, one month old and rooted
Homesaved slips or sprouts of sweet potato in April, from the previous year’s harvest

Seeds worth saving

Here I explain the different methods you need for a range of vegetables. Some are easy to save seed from, while some need more time and space, and careful selection of the seed-saving stock. If you have not saved seed before, I advise starting with any of French bean, lettuce, pea or tomato. These four vegetables neither need nor are altered by cross-pollination.


The best results happen in a dry summer.

  • First, grow your onions, in the preceding year.
  • Select the largest and best, and replant them in early spring, say, 5 cm / 2 in below surface level. The stems may need support because they grow tall, and the seedheads are quite heavy when wet.
  • Check for ripeness by rubbing a floret or two, in search of the hard black seeds.
  • Cut off the heads and put them somewhere very dry, until almost crisp, then rub out the seeds by hand, wood or foot.

Last year’s Sturon onions going into the ground to grow heads for seed

August – a harvest of onion seedheads from the polytunnel, a little immature but it was a damp summer, and mildew was developing in the tight clusters after grey mould had killed most of the leaves

Peas and beans

These are easy to save because seed is mature just one month or so after the normal period of eating them.


From mature peas at their normal picking stage, it takes two to four weeks of summer weather to have dry peas for a seed harvest, and for French beans a little longer.

Peas and French beans do not cross-pollinate, therefore:

  • Just one plant of any one variety will give viable seed, true to type.
  • Different varieties in close proximity stay true to their variety when you save the seed.
Pea pods of Oregon Sugar Pod have dried on the plants, and can now be picked for seed
Peas for seed after a damp summer, showing some discoloration of pea pods; next we shell them out
Pea seeds sown direct into pre-watered drills, after broccoli – less than 50% emerged because of dry soil; using transplants would have got a better result


These grow in the same way as peas, except that the pods turn black when dry and with ripe seeds.

The main difference from peas is that cross-pollination happens, so if you save seed from, say, two varieties in the same garden, or on an allotment site where many are growing, the new seed will not grow true. The same applies to runner (pole) beans.

One reliable variety to start with is Aquadulce Claudia.


These are easy to save but later to mature than peas and broad beans. In regions with cool, damp summers and autumns, they may not be viable. Even at Homeacres, I lose a small amount of potential seed harvest in some Octobers. They do cross-pollinate.

We shell out the Czar seeds pod by pod, but Borlotti pods are smaller and also hard when dry. This makes it more effective to walk on them, in order to crack open a lot of pods very quickly.

By August, we have saved seed of many pea varieties and of one broad bean variety
Czar runner (kidney) beans are good for seed when the pods are mostly dry and yellow
A Czar bean harvest in early October yielded 2.4 kg / 5.3 lb of dry seeds, to eat or resow
Borlotti beans are a type of French bean and grow true from one plant – we like them to eat as well as for seed; the photo shows a mass of bean pods, now drying on the plants in late September and ready to be picked soon
Walking on the dry, brittle pods of borlotti beans to crack them open
After separating the pods from seeds, I winnow in a breeze to clean the debris (see ‘Winnowing’ below); the harvest here was 3.2 kg / 7 lb of dry beans
Before putting them in jars, I lay the seeds on a tray in a sunny windowsill to ensure they are as dry as possible; these Czar and Borlotti beans are fully dry and can be stored for over a year in jars in the house – soak them overnight before cooking

Potatoes and garlic

I have saved garlic seed for two decades, and no problems yet! The only limiting factor would be if you suffered white rot, which can even be on some bought garlic seed, though this is fortunately rare.

Garlic ‘seeds’, by the way, are actually cloves of garlic which you separate from preferably large bulbs.

The constraint in saving potato seed is whether there may be either blackleg or virus. The former shows as plants prematurely dying with black, rotting stems (see ‘Disease’ section in Lesson 4 for more information), and the latter as leaves turning bright yellow. If free of those easily recognised problems, save seed of medium-sized potatoes, including any that are too green to eat.


These are fiddly and often take longer. The positive aspect is that you may harvest a large number of seeds from one plant: over 2,000, say, for lettuce.


In dry climates these are easy, but if there is too much damp weather as flowers turn to seeds in late summer, the seedheads risk rotting just before harvest. The timing goes like this:

  • Sow in either early autumn or late winter – lettuce plants survive frost, so autumn sowing works well where winters are not too cold, say above –10°C / 14°F. Plants have more time to produce seed in the following summer when they are sown in early autumn.
  • Leave unpicked in spring, so that their heart or head opens into an uprising flower stem by early summer.
  • Small yellow flowers develop in summer, into tight clusters of ten to fifteen seeds.
  • Once you can separate seeds from the clusters, twist out the whole plant and hang it somewhere dry to finish the seed-drying process.
  • Rub out the seeds and winnow in any gentle breeze, usually in early autumn.
Two March-sown lettuce, now at the stage of seed development
Late July in the polytunnel – one lettuce left to seed is ready to pull
Rubbing the seedheads with wood to separate the seeds
Grenoble Red homesaved lettuce seed, sown just 17 days earlier
Grenoble Red, Bijou and Winter Density, all from homesaved seed and 75 days since sowing


These seeds are more difficult to save than those of lettuce, because you need at least six to eight plants for cross-pollination and the seeds are clustered in florets, which are slow to mature and difficult to separate from plant stems.

Wait until the stems and seeds are reasonably dry, before the end of summer, then twist out he plants and rub off their seeds.

It’s the same method for growing beetroot seed. Plant a minimum of six beetroot in early spring; the seeds dry by late summer and you can then cut the umbels before the seeds fall out, allow them to dry, and then rub out the seeds.

Oxhella carrots for seed, stored through winter and now ready to plant; to the right we have already planted eight beetroot for roots
By September the beetroot look like this, and one root has over a thousand seeds which are easy to rub off the stems
  • Subsequent growth from the spinach in the photos, below, was variable to the point that it behaved like a different variety: the leaves were darker green, rounder and more prone to slug damage. Perhaps there were too few plants to ensure cross-pollination.
  • Medania is called a ‘type’ of spinach as opposed to a ‘variety’ or ‘cultivar’, so perhaps this variation is normal when saving its seed; we did not select it for any desired traits.
  • Subsequent seed saving in summer 2020, from these ‘new’ Medania, has resulted in stronger plants, all dark green and less troubled by slugs than normal.
  • My new strain starts more slowly, and then grows for three weeks longer before going to seed, compared to its ‘parent’ Medania.
June – 12 overwintered plants of spinach Medania for seed production
Four weeks later in July – the plants are dead and the seed is drying
Late July – removing seed from spinach stalks
The spinach seed after winnowing, still with plenty of bits
We sowed seeds on 9th August and this is 12th October, showing original Medania from CN Seeds on the left and homesaved Medania on the right


Again, you need cross-pollination and preferably an overwintered plant, since mustards are biennial, like spinach.

Sow in September, then by April you will have yellow flowers and many insects, with seed pods developing in May and ripening in June to July. We then twist out plants and walk on them to break open the small pods.


This is best done outside in a slight breeze. The idea is for the breeze to blow away fragments of pea or bean pods, lettuce fluff, bits of dry spinach leaf or small pieces of coriander pods.

  • Using two buckets, let seed drop from one bucket to the other in a steady flow, and always watch the seed going in rather than fluff flying out, to be sure that the seed itself is not blowing away. Repeat two to four times, and beware of wind gusts.

Often there are a few unwanted bits after winnowing, and you can pick out the larger ones. Most of the non-seed residue is dry and does not affect keeping quality or germination percentages. Clean the seed as much as you can, without worrying that it looks imperfect.

Walking on pods of mustard for seed, from overwintered plants
After removing the pods, I am winnowing the mustard seeds in a breeze
The pretty seeds after winnowing – this is white mustard for green manure


Be sure to only save seed from open-pollinated varieties, not hybrids (see below).

  • Tomatoes do not cross-pollinate. This allows us to save seeds from tomatoes of two different varieties, grown right next to each other.

If growing more than one plant of a variety, save seed from a fruit growing on the plant that has the nicest growth habit and harvest:

  1. Cut your tomato, as if preparing it to eat, and scoop out any seeds using a spoon.
  2. Place the seeds, and any tomato fruit sticking to them, in a cup with some water, and leave in the kitchen to ferment. This breaks down germination inhibitors on the seed, through fungal action.
  3. After five to seven days, you will probably notice a black mould of decaying fruit pulp on top of the water, which you can scoop off; the seeds will be at the bottom, now much cleaner.
  4. Give seeds a rinse and then place, say, on cardboard, in a dry place for a few days. Then pop into an envelope or an old seed packet, clearly labelled.
  5. Tomato seeds store for up to ten years, in any dry environment. A dry and ambient temperature is better than a cold and damp one.

Other vegetables

For all other vegetables not mentioned so far, I recommend that you are careful before setting out to save seed, especially in a small garden. Cabbage, carrots, kale, leeks or sweetcorn, plus many others, all need several plants growing together to ensure that there is a sufficient gene pool for healthy cross-pollination and maintenance of the varietal traits. This means that you need a lot of space to grow them. Also, root vegetables take a second year because the flower stems don’t develop until year two, after you have grown the harvest in year one.

Loss of varietal traits

A drive for profitability has resulted in less work to maintain ‘trueness to type’ of older varieties, because revenue is higher from the exciting results of new work on hybrids.

  • Maintenance of an open-pollinated (OP) variety / cultivar involves selecting the plants whose traits are most true to the characteristics desired. This is the basis of seed production for older and open-pollinated vegetable varieties such as Boltardy beetroot, Greyhound cabbage and White Lisbon spring onions.

A sad example of cost-cutting is with the Gardener’s Delight tomato. The fruits are now considerably larger, thicker-skinned, of paler colour and less sweet than they were in the 1980s, when it was the standard and best cherry tomato. This loss of quality makes F1 seed more desirable.

I have grown Greyhound cabbage for 40 years and it has always hearted early, starting in late May. However, in 2019, the plants didn’t make firm hearts at all. I had feedback of similar results from other gardeners. We waited and waited, for a disappointing and late harvest.

Gardener’s Delight in the polytunnel in August – low yielding and of poor quality
Even by 25th June, these Greyhound cabbage from Mr Fothergills have no heart

Often, if you grow enough plants of non-hybrids, you get an idea of any variety’s growth characteristics and uniformity, or lack of it. Usually there are a few, but not too many, genetic throwbacks (also called ‘sports’) to a wider gene pool, such as white beetroot and yellow carrots, which may also be tough and fibrous.

With chicories, there has been insufficient breeding in general, except by a few sellers, notably Sativa in Switzerland and Bingenheim in Germany. Until I discovered the latter two suppliers, all Palla Rossa chicories I grew, except for hybrids, were 30–40% useless for hearts, and often rotted once there was any firmness.

Avonearly beetroot from homesaved seed; it’s common to find a few white beetroot
A September harvest of Palla Rossa Romea, all from the same seed packet

I reckon the three example in the photos below are simply a result of poor maintenance. The radish Poloneza was brilliantly uniform in 2016, with a nice root-leaf balance.

Just two years later, the Poloneza I bought from the same company (Kings) had a higher proportion of leaf and less plumpness of root.

The St Valery carrots were my worst experience to date (2019) of bad seed maintenance. When I sent the photo to Mr Fothergills, they said how much they agreed with my concerns, and were themselves appalled!

One clump of Boltardy beetroot from Kings Seeds, multisown in late February and transplanted on 22nd March
The result of poor seed selection – new Poloneza radish on the left and from the original two-year-old seed packet on the right
Compare the photo on the seed packet with the harvest; plus the writing on the seed packet says ‘very uniform roots with excellent colour’

In summary

Buyer beware. Raise your own seed where you can and complain to seed companies when things go awry – it’s usually not your fault.

For the cost of new seed, it’s worth starting again with a different batch. My impression is that buying online should give fresher seed than buying from a store.

Shelling seed of broad beans, after I had walked on the crisp pods

sow & propagate
Transplant - Size, time of year, Spacing, support
container growing
Prune and train plants/thin fruit
Harvest times and method
Potential problems