Sometimes, I receive an unlabeled plant start from a friend, and they do not know if it is patented. If I want to grow and sell a plant in question, is it okay to just take the chance, or is there a way to find patents?
Some plants or shrubs can be distinguished by specific traits like leaf colour. The Diablo Ninebark with it's purple leaves is hard to miss. So if you get a purple leaved Ninebark and propagate a hundred of them for sale you would be violating the patent.
I am not a patent lawyer but I suspect that it is your intent as much action that means more. If you don't ask questions and propagate with intent to make money that will be trouble. If you receive a gift which is hard to identify and grow it in your garden it's not worth anyone's time to identify you or prosecute.
@Dmitriy Likhten Yes, farmers have been sued by agro corporations for patent infringements. However from the original question it does not seem that @jmusser will be planting a whole field of plants with intent to sell them for profit.
My knowledge of IP law is pretty limited, but from a semester of patents in law school, I can assure you that "good intentions" will not save you from a lawsuit. There is an exception for "purely scientific inquiry" if you are not going to use any research for profit (we're talking backyard invention kind of stuff here). The reality is that lawsuits are expensive and more than likely if you were "caught", you'd just receive a few cease and desist letters or be given an opportunity to setup a licensing deal with the patent holder. Licensing is where the real value of a patent is anyway. Since you're not a mega-corp (otherwise you'd have in-house counsel to direct your question at), you have no assets for a patent holder to go after anyhow.
But, you can scrap that entire explanation under the same principle a previous poster already mentioned - there is basically no way in hell anyone will discover whether you're violating a patent unless you start selling the patented plant in large quantities. At that point, you'd benefit from professional advice, and you'd have the cash to afford it. Patents themselves are confusing (almost intentionally designed that way), so there isn't really anyway for you to go do research on your own. In reality, the system is designed so that at whatever point you realize you're infringing, you'd better setup a licensing deal or stop infringing. But, there really isn't a lot of incentive to put in the leg work upfront, especially if you're small-time.
It's tough without a name, but internet image searches have gotten a lot better since the question was asked in 2012.
If you can find a name, check to see if it's trademarked (US trademark site). Easier to get than patents, they can still can get you in trouble.
In the US, plant patents are covered by the Plant Patent Act of 1930. Here's an early example of one. Here's general information (US) on what it takes to get a plant patent, and what benefits derive from having one.
Plant patents are mixed in with general patents at the USPTO search site. However, all plant patents seem to start with the letters 'PP'. The proper name of a plant, if you can find it, seems to be sufficient to ferret out any related plant patents.
There are a few issues here at least in the US, namely genetic engineering utility patents. plant patents, plant variety protections, etc. Each of these covers different aspects and we don't know what the scope of genetic engineering patents are until the Supreme Court issues their case in Bowman v. Monsanto.
I don't actually think there is a sure-fire way of telling if a plant is protected under any of the above laws.
However more to the point, if you can't tell if the plant is patented, it will be relatively difficult for the patent holder to do so either, so if it were me I would assume against patents but that is my own risk assessment for myself and may have very little to do with you. I second the idea that you should probably seek help from a lawyer specializing in intellectual property issues regarding plants because you can get a much better assessment that way, as well as a better assessment of risk.
Edit: Now that the Supreme Court has decided Bowman v. Monsanto, there still isn't much clarity. All we know is that genetic engineering patents are not exhausted through first sale when the seeds are then used to produce more plants. But then the behavior at issue in Bowman was pretty clearly intended to infringe on Monsanto's patent. Intentions aren't clearly relevant to patent infringement but I doubt Monsanto can sue just because of the probability that the soybeans one is growing probably contain some with the patented genes. If more specific analysis is required there, then it seems to me that intentions end up in the analysis through the back door.
But then other (non-GM) plant patents don't cover (in the US at least) products of sexual reproduction so you can take two patented plants and breed them and sell the offspring (plant variety protections do, but subject to some substantial limitations).
Again, my answer has not changed since I posted it, just commenting on further developments.