A friend of ours recently saw a ridiculous and completeley avoidable accident:  A car, driven at typical french speeds careened around a corner, across a crosswalk in "walk" mode, and over a womans dog (which was on a leash).  Last month a similar thing happened to me, but out of sheer luck I escaped with tire marks up my left shin and foot.  It made me think a little more about how the speed, aggressivity and carelessness of drivers in france has any effect on the rate of accidents.  As it turns out, there is a special government sponsored web site devoted to this kind of information.  I started clicking around, and interestingly, the first thing I found was a map by region of accidents/kilometer for 2000-2002.  I think the standard expectation for someone who has visited France would be that Paris would be the worst, but it turns out that its actually a small section of road near Nice, which averaged more than .8 accidents per kilometer!  When you think about that in absolute terms, and consider that the section is roughly 30 kilometers long, thats a lot of accidents; and these are only the "Serious" ones.
Moving on, there is even a PDF (in english), with a suitably macabre preface: "The international definition of "death" was amended by the Convention of Vienna in 1968: "Victims of road accidents are considered as having been killed if they die immediately or within 30 days following the accident".  I guess I'm a little naive, but it had never really occure to me that there would need to be an international definition of "death"!  Anyway, moving on:
               Deaths+Injuries                       Deaths
Country 2003    2002    Variation (%)     2003    2002    Variation (%)
Germany 354,534 362,054   2.1              6,613 6,842   3.3
Spain    99,987 98,433    1.6              5,399 5,347   1.0
Italy   225,141 237,812   5.3              6,015 6,736   10.7
UK      220,079 228,535   3.7              3,658 3,581   2.2
France   90,220 105,470   14.5             6,058 7,655   20.9
This is not a fair comparison for many reasons, the most obvious of which is that there is no normalization for number of drivers, kilometers of road etc.  More seriously, as the PDF states, the methods and obligations for reporting are very different between different countries.  However, I think the most interesting comparison is between the deaths and deaths+injuries columns.  For example, if you look at the percentages of fatal injuries in 2003, you get:
Germany   1.8
Spain     5.4
Italy     2.7
UK        1.7
France    6.7
So if you're going to get into an accident, make sure it isn't in France, since you have a strong chance of it being fatal!  It should be noted that there are several other smaller EU countries with higher percentages (like Greece).
There is another nice table with fatal accidents per million inhabitants and shows that among the big five
Germany 80.1
Spain   128.0
France  101.6
Italy   105.2
UK      61.4
which puts france in the middle of the pack in terms of deaths per capita.  Another thing i've wondered about is whether those late teen jackasses tailgating you everywhere tend to die more often than older drivers, but the data has been sliced in strange ways, so its a little hard to tell.  The most useful statistic is the deaths/million inhabitants in this group.  But this isnt perfect and is likely to be a n underestimate of the actual number since 1)"inhabitants" is not the same thing as "drivers" and 2)this ratio is not likely to be contant between age groups and 3)Their intervals for age ranges aren't constant (18-24, 25-44, 45-64, 64+.  Why is the second bin so big?).  However, here are their numbers:
Deaths/million inhabitants in this group
                  0-17 years    18-24   25-44   45-64   64+
Germany           34.4          208.0   83.4    60.8    92.0 
Spain             46.0          242.7   145.6   111.1  114.3
France            36.9          237.3   117.7   78.5   113.1
Italy   no data?
UK                26.1          147.2    71.0   44.9   69.2
So yes, that 20 year old guy passing you on the turns at 150 kmph is contributing to the very high per capita death rate (more than twice that of 35-44 year olds) in that age group.
Its impossible to tell if the teenage moped kids die with any greater frequency than other motorists (my criticism with the data is that the numer of deaths should be normalized by the number of that vehicle type, since there are many more cars than mopeds).
OK, FINALLY, the data that I was looking for:  whether you are more likely to die as a pedestrian in france.  The answer is no, at least compared to other EU countries.  However, the number is not given in per capita, but as a percentage of all accidents in that country, so I guess more searching is in order.  Also, I wonder how these stats compare to the states.