The American Association of Physics Teachers hosts an annual high school competition exam called PhysicsBowl. It's a forty-question test to be completed in 45 minutes by examinees. High schools pay a per-test entrance fee to obtain exams and enter into the competition. The exam is administered at school-site classrooms throughout the country in early April. Schools send the Scantron-type student answer sheets to a national office where they are scored and ranked.
Schools compete within about 16 regions. Rio Americano typically competes in a region that includes California, Hawaii, and US Protectorates. (The winning schools and scores in this region often come from top-shelf schools in Hawaii!). The competition is also partitioned by level: first-year physics students in one division and second-year/AP Physics students in another. A school's score is determined by the four best scores returned by that school--known only after scoring at the national office.
At Rio, we have competed every year since 1991. We have won our region and division two times and come in second place twice. We have also had individual student score recognition. All noted in our PhysicsBowl Hall of Fame. There are prizes for winning schools and top-scoring students.
I order tests and administer PhysicsBowl to my AP Physics 2 students, so we compete with the best and brightest in our region. I give my students practice tests to study and offer them extra credit based on their Physics Bowl performance. And when we win our region/division, I waive the second semester final exam for the whole class.
Generally speaking, I like the PhysicsBowl. The questions are difficult: the mean is low and the distribution is unsettlingly normal (a bell curve you wouldn't want on a criterion-referenced test taken by high-achievers).
What I don't like about PhysicsBowl is the inclusion of college topics. Developing the operational form for a difficult, nationwide competition exam is no mean feat. But the content should truly be limited to what you can find in a high school physics book and/or listed in state documents of high school physics standards. So questions about charge distribution or electric field strengths in concentric conducting shells don't need to be there.
When opened to college physics content, some truly wild questions can be brought in. But what's the point? If securing a well-distributed bell curve is the justification, why not add Lagrangians, Hamiltonians, and Bessel functions?
Clearly, it would be unfair to expect high school students to possess mastery of upper-division physics topics. I would argue that high school students should not be expected to have mastered college physics, either. High school physics they should know, and know well. But not college physics. And trust me, it's far from impossible to generate low-mean questions for all audiences in the content areas of high school physics.