Since 1994 the radon studies group at the Institut de Tècniques Energètiques (INTE) of the Universitat Politècnica de Catalunya in Barcelona, Spain, has carried out a campaign of continuous measurements of the equilibrium factor (F) and the unattached fraction (f(p)) of radon decay products at four sites which are representative of different environmental characteristics on the Mediterranean littoral of Catalonia, Spain. It has been established that these parameters vary widely, F(0.03-0.87) and f(p) (0-0.72), from one site to another and with time, according to the characteristics of the site and climate. In spite of this variation, the F and f(p) parameters are log-normally or normally distributed. The measurements of F and f(p) show that f(p) is negatively correlated to F by a log-power equation, Ln(1/f(p))=1.90[Ln(1/F)](-0.68), which can be used in all the F range, instead of the commonly used power equation f(p)=aFb suggested by Stranden and Strand and other authors, which fits well for a reduced range of F. Power and log-power equations have been introduced into a simplified dosimetric model in order to estimate the effective dose per unit radon exposure as a function of F. From the log-power equation this value is quite constant and ranged from 9 nSv per Bq m(-3) h to 12 nSv per Bq m(-3) h when F is higher than 0.15. In the case of a lower F factor, a linear function that passes through 0 fits quite well. A value of 12 nSv per Bq m(-3) h is proposed for the Mediterranean littoral of Catalonia as the best estimation.