I take around 2 minutes
How long do you spend deciding upon your moves in online chess?

I don't believe I spend enough time deciding my moves. It's one of my "chess resolutions", to take more time to think out my moves. Sometimes (more than I'd care to admit to) I move impulsively and end up paying for it.

42 hours per move, according to Chess.com.
72 hours per move in tough games.
12 hours per move in easy games, in which my opponent is clearly playing in blitz mode, and making mistakes. I realize this is wrong on my part, but I find that some well timed quick play rakes in some easy points for me.
I spend at least one hour considering what I am going to play, usually broken down as follows: 15-30 minutes considering the position and identifying candidate moves, and 30-45+minutes determining which move I will play. Oddly (and incorrectly) I spend more time on book moves, because there is always a choice of move, and I like to play through a lot of master games for at least a few of the moves before choosing which to play.
My immediate goal is to decrease my online game load, and then spend more time on each of a very few games in progress deciding which moves may be candidates, and also determining which candidate to play. I also have yet to apply a consistent thought process to every move I make, so that presents another pressing goal of mine.

If we're talking time spent actually analyzing moves, my average is probably around 30-40 minutes per move. I'll normally check my opponents move but I won't respond right away - I'll think on it subconsiously for a while, then at some point during the day I'll sit down and analyze/play 2-3 moves. I prefer to have fewer games going on, but spend more time analyzing each of those games to a deeper level.
I can spend a fair bit longer than that if I'm in a very complex situation with a lot of concrete variations.

well it varies but i do think i need to spend more time thinking about my moves and possible moves of my opponent against me.

Depends on the position. Recently I had to make a move that would decide the rest of the game, I came back to it 4 times during the day and finally made my move, I'm pretty sure I converted the game from a losing to a wining position. So in my opinion all the time it takes till you are sure it's the best move there is, even if it's a few hours... Patience is the key :)

Dependent on two factors: Complexity of the position and just how much I need/want to win this particular game.

If we're talking time spent actually analyzing moves, my average is probably around 30-40 minutes per move. I'll normally check my opponents move but I won't respond right away - I'll think on it subconsiously for a while, then at some point during the day I'll sit down and analyze/play 2-3 moves. I prefer to have fewer games going on, but spend more time analyzing each of those games to a deeper level.
I can spend a fair bit longer than that if I'm in a very complex situation with a lot of concrete variations.
I do this as well although my average can end up being quite a bit longer because I sketch the position in a notebook(which is getting filled with these now) which I take to school. Whenever I don't have something to work on instead of just sitting there I take it out and think on the next move.
Sometimes I like to look back at previous moves to remember how I got there so I also copy down the move order.
This provides me with hours worth of time when I'm just spending 10-15 minutes per look max. I then follow this up when I get home. I look back on what candidate moves I wrote and what comments I gave each one and analyze each move again to ensure that my original thoughts are still true. Sometimes I find a new candidate to try and compare that as well.
When I have extra extra time(full class periods of nothing or just lots of free time around the house) I tend to make ideas for my opponent's candidate moves as well as my replies to each one.
Worst moves I've made have tended to involve dropping a pawn by accident or were moves where life came first.

Depends on the position. Recently I had to make a move that would decide the rest of the game, I came back to it 4 times during the day and finally made my move, I'm pretty sure I converted the game from a losing to a wining position. So in my opinion all the time it takes till you are sure it's the best move there is, even if it's a few hours... Patience is the key :)
I wish I knew how to play moves that can convert losing positions to winning ones.
I'm stuck having to wait for my opponents to blunder for that to happen ... alas, I usually end up resigning or being mated. I do know how to play moves that will turn a winning position into a losing one, of course....
I guess the real question is, how long did you spend on the moves that led to that losing position?
I generally play my worst moves after either very little thought or a ton of thought.

I spent about 2 hours this morning on one move in a "I know there's a mate here somewhere" position. Only wish I had the discipline to do that on all but the most obvious moves. Sometimes I just get tired of "sweating blood" over moves and especially if my opponent replies quickly, end up playing at a rapid pace, like 2 or 3 minutes a move. I usually end up berating myself later until I remind myself that I'm doing all this just to have "fun". And yes I do think I found 3 or 4 forcing sequences to either mate or overwhelming advantage depending on my opponent's response, but I've been wrong before.

http://www.chess.com/article/view/the-number-one-reason-players-lose-chess-games-is
They move too quickly.
I don't disagree with that at all. Every really bad move I've made was from not taking enough time to analyze a position and blundering away a piece because I didn't notice an attacker.
30 seconds on average