It depends on the pair traded, how active the day is, and how much profit you are aiming for.
It also depends on if you have a really bad broker that hunts
stops.
Sometimes you set a stop too close and get taken out of a trade just before it goes your way. Other times, you'll set a stop too far away and it will just end up increasing your loss on a bad trade.
I wish I could give you some hard and fast rules. It depends so much on exact circumstances. A REALLY vague rule (which I will claim full credit for and call "Pharaoh's Rule for Stoploss" if it works and will totally deny responsibility for if it doesn't) would be to see how many minutes in the time frame of the chart you are trading (doesn't work for 1 or 5 minute charts). If you are on 15 minute charts (or shorter) and planning short trades, think 15-20 pips. If you are on 1 hour charts and hope to close in the same day, something around 60 pips may be reasonable. For longer term trades placed off a 4 hour chart, think 240-ish. Obviously, more volatile pairs (like GPBJPY) need more room than slower moving pairs (like EURGBP). Of course, really exotic pairs can have HUGE spreads, so this won't work for them.
This does make Daily and Weekly charts rather scary - keep in mind how much of your account you are risking - calculate your dollar loss if the trade slams directly into its stoploss. I just watched one of my long term trades go to almost 1000 pips negative yesterday (EEEEK!!!!), but it's turning around now. Yes, there was a stoploss (ALWAYS have a stoploss), but it was another 600 pips away.
The same goes for profit targets. Set it too close and you miss too much of a big move. Set it too far, and you'll just miss your profit target and the pair will reverse on you. Assuming you are successful with about 50% of your trades, your take profit should be significantly wider than your stoploss. Some people recommend 2:1 or 3:1. Of course, you could set it to 10:1, but you would be very lucky to succeed in 10% of your trades with a risk-reward ratio that high.
It's never a bad idea if you can take trades were significant support/resistance "protects" your stoploss and where there's no major support/resistance between your entry price and target profit - preferably support/resistance should be a short distance past your profit target.
Above all, test your theories (or mine) out on demo trades or with VERY tiny amounts of real money.