Rhinomanometry is an objective method for determining nasal patency; its reliability and relevance as an aid in defining and solving problems connected with nasal obstruction have, however, received scant attention. In the present study more than 200 subjects were submitted to rhinomanometry--most of them only by the posterior technique. In 50 of these subjects duplicate determinations of the pressure drop across the nose at the flow rate of 0.3 l/s were made within a short interval; the coefficient of variation was 20--25 per cent. The rhinomanometric values in a small group showed a day-to-day variation of 55 per cent. Because rhinomanometry allows only a moderate level of accuracy the method is unsuitable for detecting a borderline case. As the influence of the variability of the method is smaller in large materials, rhinomanometry is more suitable for comparison of groups than of individual patients.