Can Dogs Eat Tomatoes - The Plant, Not the Ripe Fruit
This page is not veterinary advice. No dose, no first aid, no feeding amounts. If your dog has been chewing a tomato plant, as opposed to eating a ripe tomato, ring your vet and say which.
That distinction is the entire page, and it is the one almost every version of this question gets backwards.
The ripe fruit is not what the toxicology literature we read is concerned with. The plant it grows on is a different proposition, and so, on a careful reading, is a green one.
Getting to that answer requires a short detour through what the reference actually says, because it does not say tomato.
What Merck covers, and what it does not
The tomato belongs to the genus Solanum, the nightshades. Merck has an entry for the genus, and it is worth quoting in full because of what it lists and what it leaves out.
The plants named are "Solanum spp. Nightshades, Jerusalem cherry, Potato, Horse nettle, Buffalo bur". The toxic principle is given as "Glycoalkaloid solanine (leaves, shoots, unripe berries)". The signs are "Acute hemorrhagic gastroenteritis, weakness, excess salivation, dyspnea, trembling, progressive paralysis, prostration, death". All animal species are described as susceptible. Read 4 August 2026.
The tomato is not on that list of common names. Neither is any tomato specific statement anywhere in what we read. So we want to be precise about what follows: Merck describes a genus that the tomato belongs to, and names several other members of it, without naming the tomato itself.
How much weight that carries. Membership of a genus is not, by itself, a toxicological argument, and we say so elsewhere on this site when other people make it. What makes it worth more than nothing here is that the toxic principle Merck names, solanine, is a glycoalkaloid characteristic of the genus, and that the parts it attaches it to, leaves, shoots and unripe berries, describe structures a tomato plant has. That is a reasonable inference and we are labelling it as an inference rather than dressing it up as a citation.
Leaves, shoots, unripe berries
Those three words are the useful part of the entry, because they are a location rather than a verdict.
A tomato plant is mostly leaves and shoots. It is a large, soft, strongly scented thing that grows at exactly the height a dog investigates, in greenhouses and grow bags and pots on patios where dogs are also present. And it produces green fruit for weeks before it produces red fruit.
Botanically a tomato is a berry, which makes Merck's phrase unripe berries an uncomfortably good fit for a green tomato. We are flagging that as our reading of a general entry, not as Merck naming green tomatoes. It is, however, the reading that would make a household act more carefully, and on a question like this we would rather err there.
WHERE THE SOURCE PUTS THE PROBLEM ripe fruit [ ] not named in the entry unripe fruit [########## ] "unripe berries" leaves [################## ] "leaves" shoots and stems [################## ] "shoots" roots [ ] not addressed Merck, Solanum spp, read 4 August 2026. Bars follow the parts the entry names. Blank does not mean safe, it means the entry is silent, which is a different thing.
The sign list is not an upset stomach
It is worth reading Merck's list slowly, because it does not describe the sort of thing people expect from a vegetable.
Acute haemorrhagic gastroenteritis is bleeding inflammation of the gut. Then weakness, excess salivation, dyspnoea meaning difficulty breathing, trembling, progressive paralysis, prostration and death. That is a neurological and cardiovascular picture sitting on top of a gut one.
Merck's houseplant table gives a parallel entry for a closely related ornamental, the Jerusalem cherry, Solanum pseudocapsicum: "solanocapsine alkaloids", producing "vomiting, hemorrhagic diarrhea, paralysis, circulatory collapse". Read 4 August 2026. The same shape, from a plant people buy for Christmas decoration.
Two Solanum entries, side by side
A map of what the entries name, not a ranking of danger. The third and fourth rows are inference and silence respectively, and both are labelled as such above.
What we are deliberately not printing
Merck's Solanum entry includes a treatment column, and it names specific pharmacological agents. We have read it and we are not reproducing any of it. There is no circumstance in which a household should be selecting a drug from a table, and printing one here would be the most dangerous thing on this page.
We are also not giving any threshold. Merck does not attach one to this entry, and the honest consequence is that nobody can tell you how many leaves matter.
The greenhouse, which is the real exposure
Most people arrive at this question because a dog ate a tomato off a plate. The more consequential version happens in a garden, and it has a seasonal shape.
Early in the season a tomato plant is all leaf and shoot with no fruit at all, which is the composition the entry describes. Side shoots get pinched out and dropped on the floor, which puts loose leafy material at ground level, exactly as pruning does with the shrubs covered on our oleander page. Trusses of green fruit appear and stay green for weeks. And at the end of the season whole plants get pulled up and left in a heap before they go to the compost.
Every one of those is a moment when the parts Merck names are more available to a dog than the ripe fruit ever is.
The potato connection
The potato is on Merck's list by name, and our potatoes page works through the same entry from the other end. The two pages are worth reading together, because the entry treats them as one problem and households treat them as two.
One point transfers directly and is worth repeating here. The popular claim that green or sprouted potatoes are the danger is not what the entry says, because the entry names leaves, shoots and unripe berries rather than tubers. We declined to assert the green tuber claim on that page and we are not asserting it here either.
What we could not source
- The tomato by name. The largest gap on the page, and the reason so much of it is labelled as inference.
- Ripe fruit. Not named as toxic, and also not affirmatively cleared. Silence, not permission.
- Any amount. No threshold in the entry.
- Timing. No onset window given.
- Cooking. Whether cooking changes anything is not addressed. Tinned tomatoes, passata and sauces were not addressed either, and a sauce usually contains onion and garlic, which are a separate and better documented problem covered on our onion page.
- Cats. The entry says all animal species are susceptible without species detail. Our cat index has no tomato specific material either.
- Rabbits. Same position. See rabbits and tomatoes for what little exists.
- Tomato pomace in commercial dog food. A common ingredient and an entirely different material from a plant. Not addressed anywhere we read.
A printable garden and kitchen check
Weighted towards the growing season, because that is where the parts the source names actually are. Prints cleanly.
Questions people actually ask
My dog ate a ripe cherry tomato off the floor.
The entry we read does not name ripe fruit and gives no reason for alarm about it. That is silence rather than clearance, so if your dog is unwell, ring. If it is bouncing around as usual, the material we read does not describe a problem.
What about a green one?
Different answer, and this is where we would ring. Merck names unripe berries among the parts carrying solanine, and a tomato is botanically a berry. We have flagged that as our reading of a general entry, but it is the reading that would lead a household to be careful, and this is not a subject where we want to be the ones talking you out of a phone call.
He eats the leaves off the plants every summer and has always been fine.
Worth telling your vet, and worth not treating as a guarantee. The entry gives no threshold, which cuts both ways: nobody can tell you what is safe, and nobody can tell you what was different about the year it stopped being fine.
Is tomato sauce a problem?
Not addressed for the tomato part, and we would be much more concerned about what else is in it. Cooked sauces reliably contain onion and garlic, and those have a well documented mechanism in dogs and cats that runs over days rather than hours. See our garlic page for how that one hides.
Why does Merck not just have a tomato entry?
We do not know, and we are not going to speculate about editorial decisions. What we can do is tell you that the gap exists, which is more than most pages on this subject will.
Related pages
Same entry, other end: potatoes. Same garden: corn and oleander. Same kitchen: onions, garlic and grapes. Indexes: can dogs eat and can my pet eat. Urgent: find an emergency vet.
How we sourced this page
Two documents, fetched and read on 4 August 2026. Treatment information appears in one of them and was deliberately not reproduced.
- Merck Veterinary Manual, Range Plants of Temperate North America, Solanum entry. Used for the list of common names, the glycoalkaloid solanine and the parts it is attached to, the sign list, and the statement that all animal species are susceptible.
- Merck Veterinary Manual, table of poisonous houseplants and ornamentals, Jerusalem cherry entry. Used for solanocapsine alkaloids and the sign list including haemorrhagic diarrhoea, paralysis and circulatory collapse.
The one thing to carry away. When you ring, say which part. Leaf, stem, green fruit or ripe fruit are four different sentences, and the source we read treats them differently.
