Stellar evolutionary models are often used to derive relevant properties of globular star clusters and galaxies, such as their age and metal content, hence their reliability must be checked. The Luminosity Function (LF) of the stellar sequences in the CMDs, from the MS Turn Off (TO) up to the termination of the Asymptotic Giant Branch (AGB), has been recognized to be the most powerful instrument for testing stellar evolutionary models (in particular the accuracy of the input physics, the reliability of canonical assumptions, etc.).
A fully fruitful test requires that the observations be a) complete, b) statistically significant, and c) accurate and adequate for each specific evolutionary sequence. Point (a) means that virtually all the stars in a given area of the cluster are measured down to a given magnitude level, and that reliable corrections for incompleteness can be applied below that level. Point (b) means that observations should cover most of the cluster extension. Point (c) requires e.g. IR observations to measure the cool RGB stars and UV observations to properly study the blue sequences as the Horizontal Branch and the Blue Stragglers.
A few years ago, we started an ambitious project aimed to a new global approach to the test of theoretical sequences [in collaboration with Paltrinieri (Univ. Roma) and Rood (Univ. Virginia)]: the immediate objective of this project was the construction of a new generation of LFs based on multi-band (from the near-IR to the far UV) observations for selected Galactic globular clusters (GGCs), in which all post-MS stars at all radii have been measured.
The validity of this approach has been shown by our work on M3: we have constructed, using photographic and CCD photometry, and HST data, the most complete CMD in a Galactic GC, covering the entire cluster from the very center to a radial distance of about 6 arcmin. This huge work has been the subject of seven papers, in the last of which [Rood et al. (1999, ApJ, 523, 752)] we have presented the global LF for RGB-SGB-MS, showing a substantial agreement of the data with standard theoretical models. Such a data set has been also obtained for other clusters with different structural parameters and metallicity (M92, NGC2808, NGC288).
The potential use of such a database is huge: the behavior of any evolutionary phase can be studied in the most appropriate photometric band allowing a high resolution analysis of the "fine structure" of each sequence.
The study of the UV sequences (e.g. the Horizontal Branch) has been addressed mainly using UV-HST observations. A large data set has been secured during HST Cycles 4-5-6-9 and the first set of data analyzed has produced a variety of interesting results.
The systematic study of the red sequences (Red Giant Branch -
RGB-AGB) has been performed using IR and optical data. IR data have
been obtained at the ESO-MPI 2.2m telescope using IRAC2, and a new
data set has been secured with SOFI at NTT and ARNICA/NICS at TNG.
In total we obtained J,K images of a sample of 
30 GGCs. The
RGB main features (colors and LF) for a first sample of 10 GGCs in the
IR and 61 in the optical have been published [Ferraro et al. (2000,
ApJ, 119, 1282)],
showing an
excellent overall agreement between the observations and the most
updated theoretical models.
We are also analyzing 10 
m ISOCAM observations of 6 massive GGCs
in order to study the mass-loss processes along the RGB with varying
cluster metallicity.