My first post!! I am so excited!!
I have seen the Mythbuster episode where they showed that 600 Volts from an electrified subway line is enough to kill a person. I also know that it is not the voltage that kills but the current. In this case, about 0.65 amps.
What I want to know is why larger voltages don’t nearly come close to killing you.
A Van de Graaf generator can produce thousands of volts but my students find it fun to touch. If the voltage is higher and the electrical resistance of the body hasn’t changed, then why isn’t the generator pushing enough electrons through your body to kill you?
More current means more energy means more power. A Van De Graaf generator simply cannot provide enough wattage to kill you. You can pretty much generate an arbitrary voltage, but you can’t just generate arbitrary current because that would require certain specific properties in the circuit (i.e. voltage and resistance)
To me, the saying ‘it’s not the voltage that kills, it’s the current’ is awfully misunderstood. It’s a bit like saying ‘its not the fall that kills, its the landing’. Since you can’t have one without the other, it’s a little silly to blame one component.
Basically, the current requires a certain voltage to reach the heart. You can 100% safely touch the terminals of a car battery that may provide hundreds of amps bucause of the low voltage. Similarly, static electricity discharges (and Van de Graaf generators) measure in the thousands of volts but carry tiny currents. It’s further complicated by the path the current travels. To kill, the current path has to go through the heart (or be backed by enough power to fry other organs).
To expand on what groman posted, all voltage sources have a finite resistance associated with them, called source resistance (see Source Resistance). This source resistance limits the current you can draw from a voltage source, no matter what you put across it. A 10,000 volt Van der Graff generator can only supply microamps of current no matter what the load (e.g. your body), due to a very high source resistance. However, 120V AC voltage from your house wiring can kill you since it has a very low source resistance. The 7200 volt distribution line outside your house is even more dangerous - the source resistance is so low that thousands of amps can flow, enough to blow things up.
Batteries don’t have enough voltage to shock you due to your body resistance - it’s high enough that it takes 30V or more to start causing trouble. A 12V car battery can certainly produce a lot of current, though.
Arjuna34
There are two ways that electricity generally kills a person.
First, you can screw up your heartbeat. It takes a surprisingly small amount of current to do this, which is where the mythbusters number of 0.65 amps probably comes from. The current “safe” value that you can pass across your heart is thought to be 0.005 amps (5 mA), but for obvious reasons we haven’t tested this out too much on humans. Killing you by screwing up your heartbeat isn’t at all guaranteed, though. You almost have to apply the electricity at just the right moment during the heart’s cycle. However, if you do manage to hit it just right, the way your heart is designed, it tends not to go back into a normal heartbeat all by itself. In other words, if someone isn’t standing next to you with a portable defibrillator at the time, you’re probably in a big heap o’ trouble. Make sure your organ donor card is up to date.
The second way that electricity kills you is by literally cooking you to death. This is how the electric chair does its thing. If you were to take two electrodes and put them on either side of a hot dog, you’d be surprised at how fast the little bugger heats up. In less than a minute you’ve got a fully cooked lunch. However, don’t try this at home kids, because two bare electrodes at the end of a wire is what us folks who deal with electricity all the time call a suicide cord. It makes for a good Mr. Wizard demonstration, though.
The funny thing is, at very low currents, it’s easy to throw your heartbeat out of whack, and at very high currents, it’s easy to cook you to death, but in between these ranges, it’s likely that you won’t die. Instead, what happens is that your heart tends to get “stuck” while you are getting shocked (all the muscles just contract and stay there so it’s no longer pumping blood), but your heart is very likely to start beating again once the electricity is removed (i.e. someone grabs a wooden broom and knocks you off of the wire).
A Van De Graff generator can be dangerous, but only if the dome is big enough. Otherwise, it doesn’t store enough energy to screw up your heartbeat. If you got a Van De Graff generator with a really big dome, though, you could potentially (oops, pun, sorry) kill someone with it.
The human body can store quite a bit of charge too, but again, there’s not enough energy stored to be dangerous. It is very common to get shocked from static electricity, especially in the winter. If you can actually see the blue spark, then the voltage is actually up in the several tens of thousands of volts.
Lightning bolts, by the way, generally kill you by cooking you to death. Here’s a few numbers to put them all in perspective. Typical electrical outlet in the US: 120 volts, limited to 15 amps max. Typical electric chair: a couple thousand volts, at somewhere around 10 amps (varies a bit from chair to chair). A typical lightning bolt is a few million volts with a few hundred thousand amps of current. You can think of a lightning bolt as essentially a really really big bug zapper. Unfortunatey, we’re the bugs.
Speaking of Van de Graff generators, I visited the Museum of Science in Boston last week and got a chance to see this monster in operation. These domes are each 15 feet in diameter. Does that qualify as “big enough”? 
Here’s the deal:
When someone says a Van De Graff generator has 10,000 volts on it (or whatever), they’re right. But when you touch it, you do not have 10,000 volts directly applied across your body. We don’t need to make fancy measurements to prove this is true; if 10,000 volts were truly placed across your body, a lethal level of current would likely flow through your body, and you would probably not live to tell about it.
As stated by Arjuna34, the secret lies in the Thevenin equivalent source resistance of the voltage source. Let’s look at the Van De Graff generator… when there is no load on the voltage source (i.e. when no one is touching it), the voltage is indeed very high. But as soon as someone touches it, the resistance of the load (i.e. the resistance of the person’s body) and the Thevenin equivalent source resistance of the voltage source form a voltage divider. The latter is much less than the former, hence the actual voltage across the load is very low (and thus the current through the load is low).
Oh, and for the record… current is what actually kills you, not voltage. But it takes voltage to produce the current. The voltage required to kill you will depend on your resistance. 30 V might kill you if it’s applied directly to your blood vessels, and 150 V might not kill you if it’s applied to very dry skin.
You know, drifting off topic, I tried this once and I didn’t end up with a fully cooked lunch. I ended up with a mostly-raw hot dog that had 2 burnt ends. Maybe it was because my source was an AC wall outlet.
Anything above 50V should be treated with caution, though as mentioned if the energy available is low then even high voltages can be touched, though as mentioned again they cease to be high voltages soon after being crowbarred by a relatively low-resistance fleshy body.
IIRC, school physics lab van der Graaf generators have quite a low source resistance, but the energy is limited by the small capacitance of the charge collector dome. The anode cap on a TV cathode ray tube can also hold a large voltage on its capacitive plates, but here the capacitance is much larger, and stores enough energy to kill. Another high voltage source - the domestic room ioniser - has at least one big resistor (between 10Mohm - 100Mohm) limiting the current before it reaches the external emitter. Built-in electronic gas cooker ignitors might generate thousands of volts, but the energy of each spark is limited by the relevant mandatory design standards, and is tested with a simulated finger. As with the v.d. Graaf generator the source resistance is low, and though the initial current through the body can be quite high when touched, it’s for a very short duration, and the total energy is limited.
RCD safety trips will turn off the mains after 30ms of a 30mA shock, so I guess it’s assumed that most people will survive these figures.
I might be missing something, but whenever I’ve seen a Van De Graff generator demonstrated the person touching it is usually insulated from anything else. In short, they can’t conduct any current because there is no potential difference across them. They might be at 10,000v relative to the Earth, but since there isn’t a current path they are just fine.
Technically, you can grab a 100,000v line as long as you aren’t grounded. Potential is relative.
Just to let you all know that way back when I started posting I made the statement that “it’s the amps, not the voltage” and got my ass handed to me pretty damned quick by Q.E.D. But please, carry on.
Well, you were right… it is the current. Opinions vary, but it is generally agreed that the “fatal threshold” is around 60 to 70 mA of current through the body.
As mentioned by me and others, you need a voltage to produce a lethal current. The formula for the voltage is:
V (volts) > 0.06 * total body resistance (ohms)
Easy formula, right? Right. But the hard question is this: what’s the total body resistance? 
Answer? It depends. If, for example, the entry and exit points are through broken skin, the body resistance might be pretty low (like around 800 ohms), which means it won’t take much voltage to shove 60 mA through the body (around 48 V). On the other hand, if the entry and exit points are made on very dry, clean, and thick skin, the resistance might be pretty high (like around 10,000 ohms), which means it would take a relatively high voltage to shove 60 mA through the body (around 600 V).
My numbers may not be realistic. But you get the drift…
There’s often a second part of the demonstration where that person reaches out and touches someone who is not insulated, resulting in a spark and shock. This is safe to do with a small Van de Graff generator because it’s essentially a very small capacitor being charged by a very low-current source.
Just for the hell of it, I went back to 2003 and found that thread. It wasn’t Q.E.D. who pinned my ears back, it was…ready?..Crafter_Man! 
There’s a big difference between DC and AC power in terms of safety. To my understanding, AC power (usually) kills you by screwing up your heartbeat, which only takes a small amount of current (though it must go through the heart itself). DC power kills you by cooking you, which requires significantly larger currents.
DC can screw up your heartbeat as well, but it is most likely to do so when the DC is first applied to your body (in other words when you get transient amounts of current flowing and it hasn’t settled down to DC yet). However, if you want to stop a heart, one of the best frequency ranges to do it is around 50 or 60 Hz. We almost couldn’t have picked a worse frequency to use from a safety standpoint. AC is more likely to stop your heart at low current levels, but DC isn’t safe.
At high current levels, AC and DC both cook equally well.
Glad to be of service. And I’m still keeping an eye on you, so straighten up!
I think it’s safer (from a pedagogal point of view, at least), to just point out that voltage, current, and resistance are inextricably tied together, and leave it at that. You can’t blame the voltage or the current separately unless you could change one variable without changing the other two, and you can’t do that.
I’m confused, however, cuz back then you told me (rather irritably) that it was indeed the voltage and not the amperage. In this thread, you seem to contradict that. S’plain?
Chefguy:
I know it seems confusing, but my position hasn’t changed.
In that thread you said,
I was focusing more on the “90v vs. 120v is irrelevant” comment than anything else. I pointed out that voltage is important in the sense that you need enough voltage to produce the lethal amount of current.
Let me try again.
It’s all based on Ohm’s Law:
V = IR
where
V = Voltage across the body (volts)
I = current through the body (amps)
R = resistance of the body (Ohms)
Medical researchers have “calibrated” the lethality of electricity in terms of I (current through the body). They have determined that the risk of death greatly increases when the current is above 0.06 amps.
So here’s the question: How do you get the current to be greater than 0.06 amps through the body?
Answer: V/R > 0.06
If V/R > 0.06, then the risk of death greatly increases.
Based on the equation V/R > 0.06,
If R is very high (e.g. current enters & exits through thick, clean, dry skin), then you need quite a bit of voltage for electrocution.
If R is fairly low (e.g. current enters & exits through broken skin), then a low voltage can cause electrocution.
Does that clear it up?