We construct a complete, semi-empirical donor sequence for CVs with orbital periods less than 6 hrs. All key physical and photometric parameters of CV secondaries (along with their spectral types) are given as a function of P_orb along this sequence. The main observational basis for our donor sequence is an empirical mass-radius relation for CV secondaries. We present an optimal estimate for this relation that ensures consistency with the observed locations of the period gap and the period minimum. We also present new determinations of these periods, finding P_{gap, upper edge} = 3.18 +/- 0.04 hr, P_{gap, lower edge} = 2.15 +/- 0.03 hr and P_{min} = 76.2 +/- 1.0 min. We then test the donor sequence by comparing observed and predicted spectral types (SpTs) as a function of orbital period. To this end, we update the SpT compilation of Beuermann et al. and show explicitly that CV donors have later SpTs than main sequence (MS) stars at all orbital periods. The semi-empirical donor sequence matches the observed SpTs very well, implying that the empirical M2-R2 relation predicts just the right amount of radius expansion. We finally apply the donor sequence to the problem of distance estimation. Based on a sample of 22 CVs with trigonometric parallaxes, we show that the donor sequence correctly traces the envelope of the observed M_{JHK}-P_{orb} distribution. Thus robust lower limits on distances can be obtained from single-epoch infrared observations.