CIE A-Level Physics Paper 3: The practical paper explained

If you've been looking for a list of required practicals for CIE A-Level Physics (9702), you won't find one. Unlike Edexcel (16 core practicals) or AQA (12 required practicals), Cambridge International assesses practical work by testing a defined set of skills in unfamiliar contexts, not a numbered recipe book of experiments. The paper you walk into a lab for is Paper 3, the Advanced Practical Skills paper. Its sister paper, Paper 5, is a written planning and analysis exam with no apparatus.

Paper 3 has a specific structure that most revision guides underplay. It consists of two experiments, each worth 20 marks, each lasting about an hour. The two experiments are set in different areas of physics, so you might go from a mechanics experiment to a circuit experiment on the same paper. And here's the detail that catches people out: Question 2 is deliberately designed to use an inaccurate method. The point of Question 2 isn't to get a perfect reading, it's to evaluate what's wrong with the method and suggest improvements.

This guide walks through what the two questions actually look like, the apparatus you need to be comfortable with, how the 40 marks split across skill areas, and the specific trap that costs students time on Question 2 every year.

The headline information you need to anchor your revision, with Paper 3 and Paper 5 compared side by side.

The 9702 syllabus is explicit that no prior knowledge of the theory will be required for Paper 3. That's reassuring, and it's genuinely true: the physics context in the exam may be entirely unfamiliar. What the markers are checking is whether you can set up apparatus, take readings to the right precision, plot a graph cleanly, and evaluate your results. The content is a vehicle for the skills.

Question 1 is the "standard" practical. You're given a method, you collect data by varying one quantity and measuring another, you tabulate the results, you plot a graph, and you draw a conclusion from the gradient or intercept. A typical Question 1 might ask you to vary the length of a wire and measure how its resistance changes, or vary the mass hanging from a spring and measure the oscillation period.

The mark split for Question 1 (minimums): manipulation and measurement 7 marks, presentation of table and graph 6 marks, analysis and conclusions 4 marks, with the remaining 3 marks distributed depending on the specific experiment. What that split tells you is that doing the practical cleanly (collecting good data) carries more weight than the final calculation. Students who rush the readings and then try to pull a tidy graph out of messy data lose more marks than students who take their time on measurement and get a straightforward graph.

Question 2 is where candidates most often lose time. The 9702 syllabus says it plainly: the experimental method to be followed will be inaccurate, and candidates will be required to evaluate the method and suggest improvements. The method you're given has known weaknesses baked in: perhaps you can only take two readings instead of a range, or the apparatus has a resolution that's too coarse for the quantity you're measuring, or there's a systematic error (parallax, uneven heating, friction) that you're expected to notice.

The trap is that students see the flaw and try to fix it before taking any readings. They set up a cleaner version, lose 10 minutes to the extra set-up, and then run out of time on the evaluation. The point of Question 2 is not to produce a perfect result. The point is to run the method as given, record the data you get, then use the data (and your observations) as evidence for your evaluation.

The mark split for Question 2 (minimums): manipulation and measurement 5 marks, presentation 2 marks, analysis, conclusions and evaluation 10 marks, with 3 marks distributed. Half the marks on Question 2 are in the evaluation section. If you don't get to it, you can't score them.

The 9702 syllabus publishes an apparatus list, and anything on it can appear in Paper 3. If you've only ever seen some of these demonstrated, your hands will be slower than your brain on the day. Try to use each one a few times.

Marks are lost (and gained) on precision alongside raw accuracy. Each piece of apparatus has a resolution, and your recording should match it:

Micrometer: 0.01 mm. Record 2.47 mm, not 2.5 or 2.470.

Vernier calipers: 0.1 mm. Record 24.3 mm, not 24.30 (that implies micrometer precision).

Stop-watch: 0.1 s. Record 12.3 s, not 12.30. If you're timing oscillations, time 20 of them and divide, so a 0.2 s human reaction error becomes 0.01 s per oscillation.

Thermometer (–10 to +110 °C, reading to 1 °C): record to the precision of the scale. Record 24 °C, not 24.0 or 24.50.

Digital meters: match the display. If the display reads 1.23 V, record 1.23 V, not 1.2 or 1.230.

These look like trivial details, but they come up on both Paper 3 and in the "indirect practical skills" questions on Papers 1 and 2.

A clean graph is worth multiple marks, and students who rush straight to the gradient leave easy marks behind. Here's what Cambridge expects, in order of importance:

Axes labelled with the quantity and unit, for example "length / cm" or "period² / s²". The slash notation (quantity / unit) is Cambridge's preferred format.

Scales that use the graph paper. If your plotted points cover less than half the grid, you've chosen scales that are too wide. Pick scales so the data fills at least half the paper in both directions.

Points plotted as neat crosses or dots in circles, not oversized blobs. A sharp pencil helps.

A best-fit line drawn with a ruler, continuous, with roughly equal numbers of points above and below.

The gradient calculated from two points well-separated on the line (not two of your plotted data points), with the coordinates read off and the calculation shown. The y-intercept read off cleanly if the x-axis starts at zero.

If you're asked to find a quantity from the gradient (for example, finding resistance from a V/I graph), state the equation, substitute the gradient value, and give the final answer with units.

Three mistakes come up every year, and all three are preventable with lab practice rather than content revision.

First, trying to fix Question 2's method before taking readings. Already covered above, but it's among the biggest time sinks. Follow the method, record the data, evaluate afterwards.

Second, false precision. Writing 24.30 cm when you measured with a 1 mm ruler is wrong, and it's picked up every year. Write what your apparatus can read, no more.

Third, vague improvement suggestions. If you say "use a more accurate ruler" without naming the apparatus or the resolution improvement, you get nothing. Say "use digital vernier calipers reading to 0.01 mm instead of a 1 mm ruler, so the uncertainty in each reading drops from ±0.5 mm to ±0.005 mm". That's a specific, named improvement linked to the weakness.

Unlike the reformed UK A Levels, Cambridge International 9702 has no Practical Endorsement, even if you sit it at a UK centre. There's no separate teacher-assessed pass or fail: your practical skills are assessed entirely through Paper 3 and Paper 5, and both count directly towards your grade.

You can't prepare for a practical exam by reading methods. You need lab time and you need to practise against the specific skill areas Cambridge tests.

Start with past Paper 3 papers from the Cambridge website. Walk through each experiment on paper first: What apparatus would you use? What readings would you take? What would the table look like? Then compare to the mark scheme and see where your answer differs. For broader exam technique beyond the practical paper, see our guide to getting an A* in A-Level Physics.

Next, do lab time. Specifically: take 20 readings of a wire's diameter with a micrometer and work out how much they vary. Time 20 oscillations of a pendulum and divide by 20. Build a potential divider circuit from scratch and measure the voltage across a component as you change the slider position. Plot the results by hand on graph paper. These are the skills Cambridge tests, repeatedly.

Finally, drill the "evaluate this method" skill. Take any past Question 2, read the method, and list every possible weakness you can think of (small range of readings, large uncertainty, systematic error, difficulty keeping a control variable constant). Then write a specific improvement for each weakness. This is the half of Question 2 you can't fake in the exam.

Our CIE A-Level Physics course covers the full 9702 syllabus with worked examples of the apparatus, calculations and evaluations Paper 3 tests. For broader revision strategies, our A-Level revision techniques guide covers active recall and spaced practice.