With all the recent discussions around here about safety, I think that there's one thing that all of us working chemists can agree on: MSDSs are often the next thing to useless.
They're not supposed to be, at least in theory. The idea is that a materials safety data sheet collects all the relevant toxicity, handling, and disposal information for a given chemical so it can be referenced by users, emergency responders, and so on. But somewhere along the line, things have gone well off track. I refer interested readers to the famous example of the MSDS for sand. Sea sand.
The first thing we find is that it is a cancer hazard. Then we note that "Prolonged exposure to respirable crystalline quartz may cause delayed lung injury/fibrosis (silicosis)". Which is true, but (of course), we have no idea of what "prolonged" means in this context, and we may not realize that sand, in its commonly encountered forms, is not easy to inhale. One should " Wear appropriate protective clothing to prevent skin exposure", but if we were to contact this substance through our own carelessness? We should "Immediately flush skin with plenty of water for at least 15 minutes while removing contaminated clothing and shoes.". We should take care at all times: "Do not let this chemical enter the environment." But that should go without saying, since we've been enjoined to "Use only in a chemical fume hood".
Now, what this thing is trying to tell us is that extensive exposure to finely ground silica dust is bad for the lungs. This is absolutely true, even if lawyers have been trying to make dubious fortunes off of it. A person should take care not to inhale sand dust, and should take particular care if exposure to such dust is a regular feature of one's job.
But there needs to be a way to get this information across without making a bag of sand sound like a weapon of mass destruction. I don't know how many times I've heard of chemical spills being treated like high-level radioactive waste because emergency responders (or local news reporters) read the MSDS and hit the panic button. (A famous example was the closure of the Bay Bridge in California once by a few bags of iron oxide (keep in mind that this happened before the current environment of worries about terrorist incidents). The responders knew what the chemical was: they read the MSDS, which (naturally) told them to wear full protective equipment, avoid exposure, wash copiously and seek medical attention, etc. For a few bags of rust.
There has to be a better way - you'd think, at any rate. But the MSDS is lawyer language, when you get right down to it, and there's the problem. Trying to insulate everyone from liability is not something that can be done simultaneously with trying to inform people in case of an emergency. Very few chemists, in my experience, spend much time with these forms at all, preferring to get their information from almost any other source. There has to be a better way.