I can’t think of any natural foods that are sweet and poisonous. Is sweetness always an indicator of no poison in natural foods?
The Tree TutuCorairia arborea of New Zealand has sweet berries that can cause poisoning if swallowed. However, the poisonous part is the seed. The juice can be used if the seeds are strained out.
Unripe Ackee is pretty dangerous
Also apple seeds contain amygdalin, a compound that produces cyanide. Although humans can safely digest small doses of plant cyanide, larger doses can lead to dizziness and vomiting, among other symptoms. Severe poisonings lead to increased blood pressure and heart rate, kidney failure, and coma. The ensuing respiratory arrest can lead to death.
I am one of those weirdos who eats the whole apple, steam, seeds, core, and all. So I have probably ingested more than the average consumption of amygdalin and I am fine. I think the seeds are sweet.
It would be counter-productive, evolutionarily.
Plants are bitter because that discourages predators from eating them. Plants are poisonous because that also discourages predators from eating them (at least, not more than once).
Plants are sweet to encourage predators (or pollinators).
It would make no sense for a plant to be both sweet and poisonous, and such a plant would be at an evolutionary disadvantage, and likely to go extinct.
Many types of elderberry are poisonous unless cooked.
Atropa belladonna (deadly nightshade) berries are reportedly sweet, but are poisonous.
Manchineel (‘Beach Apple’) fruits are delicious, but highly toxic.
Many deadly poisonous species fungi - including the most dangerous species of Amanita (Death caps, etc) are actually tasty to eat.
Obviously false, since many counterexamples exist.
Seems logical, but the fruit might be poisonous for us and not so for any number of other species (birds, etc…)
Indeed - birds are able to eat many things that are toxic to mammals because of their rapid digestive transit - the bird digests the fleshy part of the fruit and the (often toxic) seed passes through, deposited somewhere else with a little drop of fertiliser to aid growth. Some seeds germinate much better once they’ve passed through a bird.
In other cases, toxins that affect the metabolism of some animals have no effect on others - and in yet other cases, animals offset some of the toxicity of their food by ingesting clay.
One wonders about a reproductive tactic that creates a delicious fruit, which germinates and flourishes in the rotting corpse of the animal it poisons.
In general killing your host isn’t a good long term strategy for a species, so that is probably why we don’t see it (assuming we don’t).
Finger Cherry. Sweet and poisonous.
Quite possible.
For example, many species of South American parrot don’t perceive capsaicin as a burning thing. My conure eats hot peppers like they were candy because as far as he’s concerned they are - he perceives them as sweet, not painful. That hot stuff is there to discourage mammals from eating their fruit. The plants “want” their seeds distributed in bird droppings, not digested by mammals.
Then there’s chocolate. People are one of the few critters that can safely eat it in significant quantities. It’s a lot more toxic for other animals than it is for us.
Water hemlock (Cicuta virosa), possibly the deadliest plant in North America and Northern Europe, reportedly tastes good.
Yes, but naturally chocolate is super bitter. You have to add sweeteners to make it sweet.
This is also suggested as the evolutionary advantage of capsaicin.
So is the overripe Ackee. But I never considered ackee to be sweet.
Then again, it’s been years since I’ve had any. I mostly let my backyard squirrels eat it, when they’re not…never mind. Wrong thread for that. 
What a coincidence this question should pop up. My SO came back from a Weight Watchers meeting this week and reported she’d just learned a new factoid – the reason we like sweet things is that we’re evolutionarily designed to recognize that they’re not poisonous. The “sweet” flavor cannot disguise poison, the meeting leader said.
I immediately called bullshit on that. Just the number of compounds that can be considered “poisonous” makes that claim outrageous.
However, the only counterexample I could come up with on the spot was lead. Metallic lead has a sweet taste, and is definitely poisonous, though it’s not common in nature, so probably wouldn’t exert significant evolutionary pressure.
The reason we like sweet things is a lot more obvious than that: in nature, such foods usually contain sugar, which is a great source of energy. You don’t need any other evolutionary reason than that.
Ask him why we like salt, sour, and umami tastes.
I think chocolate was an example of something not poisonous to us, but poisonous to other animals which was contrary to the examples given. Not really about the sweet factor.