We re-discuss in detail the use of linear fit to the (B - V), (B - I)
ZAMS of open clusters proposed by Natali et al, (1994) to measure the
reddening. A proper link to the Cousins photometric system and evaluat
ion of the adopted extinction law shows that the application of the pr
oposed method must be carefully handled over observational data to avo
id large errors. Other combinations of colors are explored in search f
or a higher sensitivity to reddening, which in any case remains quite
poor.