An informative site is UK Polling Report. They have threads for each (UK) European region, and some reasonable comments (ie not just "my party will smash you all!!!" chest-beating).

It looks like the most likely place for the BNP to win a seat is North West England (so that's where they've stuck Griffin, of course - the Maximum Leader gets first dibs on a seat). Thread here: http://ukpollingreport.co.uk/guide/north-w...

Under the d'Hondt system, for a party to win its nth seat, it must have n times the number of votes of a party that hasn't yet won a seat.

So, for instance, for a party to win 3 seats, it must get at least 3 times the number of votes that a party that doesn't get a seat got. NW England has 8 seats, and if the votes went:

Con 30% (+6 from 2004)
Lab 23% (-4)
LD 15% (-1)
UKIP 8% (-4)
BNP 8% (+2)
Green 7% (+1)
Others 9% split various ways
then it's Con 3 seats, Lab 2, LD 1, UKIP 1, BNP 1

To stop Griffin from that position, 2 out of 5 parties must hit their 'targets':
Tories: get more than 4 times the BNP vote
Labour: get more than 3 times the BNP
Lib Dems: get more than 2 times the BNP
UKIP and Green: beat the BNP

The Greens are calling for tactical voting for them to stop the BNP. Whether that's the really the most likely way to stop them, I can't tell - anyone think my guesses for swings from 2004 are realistic/unrealistic?

London looks less likely for the BNP (8 seats again, but the BNP got just 4% there last time); Yorkshire & Humberside is probably the BNP's next target (6 seats, and they got 8% there last time). South East England has 10 seats, but the BNP only got 3% there last time - but if there's a mass defection from UKIP to them, it could happen there too, I suppose (and that might go for East Midlands too, where UKIP got a huge vote last time).