Keeping in view of the modern powerful observing tools, among others Hinode (formerly SOLAR-B), STEREO and Frequency-Agile Solar Radiotelescope, and sophisticated modelling techniques, Joint Discussion 3 during the IAU General Assembly 2006 focused on the properties of magnetic field of solar active regions starting in deep interior of the Sun, from where they buoyantly rise to the coronal heights where the site of most explosive events are located. Intimately related with the active regions, the origin and evolution of the magnetic field of quiet Sun, the large scale chromospheric structures were also the focal point of the Joint Discussion. The theoretical modelling of the generation and dynamics of magnetic field in solar convective zone show that the interaction of the magnetic field with the Coriolis force and helical turbulent convection results in the tilts and twists in the emerging flux. In the photosphere, some of these fluxes appear in sunspots with field strengths up to about 6100 G. Spectro-polarimetric measurements reveal that the line of sight velocities and magnetic field of these locations are found to be uncombed and depend on depth in the atmosphere and exhibit gradients or discontinuities. The inclined magnetic fields beyond penumbra appear as moving magnetic features that do not rise above upper photospheric heights. As the flux rises, the solar chromosphere is the most immediate and intermediary layer where competitive magnetic forces begin to dominate their thermodynamic counterparts. The magnetic field at these heights is now measured using several diagnostic lines such as Ca II 854.2 nm, H I 656.3 nm, and He I 1083.0 nm. The radio observations show that the coronal magnetic field of post flare loops are of the order of 30 G, which might represent the force-free magnetic state of active region in the corona. The temperatures at these coronal heights, derived from the line widths, are in the range from 2.4 to 3.7 million degree. The same line profile measurements indicate the existence of asymmetric flows in the corona. The theoretical extrapolation of photospheric field into coronal heights and their comparison with the observations show that there exists a complex topology with separatrices associated to coronal null points. The interaction of these structures often lead to flares and coronal mass ejections. The current MHD modelling of active region field shows that for coronal mass ejection both local active region magnetic field and global magnetic field due to the surrounding magnetic flux are important. Here, we present an extended summary of the papers presented in Joint Discussion 03 and open questions related to the solar magnetic field that are likely to be the prime issue with the modern observing facilities such as Hinode and STEREO missions.