We present a new approach to tackle the issue of radio loudness in quasars. We constrain a (simple) prescription for the intrinsic distribution of radio-to-optical ratios by comparing properties of Monte Carlo simulated samples with those of observed optically selected quasars. We find strong evidence for a dependence of the radio luminosity on the optical one, even though with a large scatter. The dependence of the fraction of radio loud quasars on apparent and absolute optical magnitudes results in a selection effect related to the radio and optical limits of current surveys. The intrinsic distribution of the radio-to-optical ratios shows a peak at $R^*_{1.4} \sim 0.3$, with only $\simlt 5$ per cent of objects being included in a high $R^*_{1.4}$ tail which identifies the radio loud regime. No lack or deficit of sources -- but only a steep transition region -- is present between the radio loud and radio quiet populations at any $R^*_{1.4}$. We briefly discuss possible origins for this behaviour (e.g. absence of jets in radio quiet sources, large range of radiative radio efficiency, different life-times for the accretion and jet ejection phenomena, ...).