The idea is to measure the capacitor voltage with a device that has a high enough resistance that the capacitor voltage decays slowly, relative to the time it takes you to see the result. If one end of the capacitor can be grounded, you can probably use a divide by 10 (10 meg impedance) scope probe and your scope to check its voltage, and still have a small enough capacitance that it doesn't suck the peak voltage down too much as it charges. This all depends on what is producing the peak voltage, and how fragile the source is.
Something like 1000 pF may be large enough since this has a 10 ms time constant with the probe. Even without a storage scope, you should be able to see how far the trace moves in that time. If the peak occurs much more often than every 10 ms, the capacitor will have only a little ripple voltage and will absorb only a little charge on each peak.
Make sure your diode, capacitor, probe and scope are rated for the expected voltage.