It is now well-known that the surface magnetic fields observed in
cool, lower-mass stars on the main sequence (MS) are generated by
dynamos operating in their convective envelopes. However, higher-mass
stars (above 1.5 Msun) pass their MS lives with a small convective
core and a largely radiative envelope. Remarkably, notwithstanding
the absence of energetically-important envelope convection, we
observe very strong (from 300 G to 30 kG) and organised (mainly
dipolar) magnetic fields in a few percent of the A and B-type stars
on the MS, the origin of which is not well understood. A variety of
theories have been proposed to explain the existence of these fields,
but most are unable to self-consistently explain the fields strengths
and topologies, ages, and rotational properties of these magnetic
stars. Currently, the leading model is the so-called "fossil field
hypothesis" which proposes that these fields are the slowly-decaying
remnants of interstellar magnetic field, swept up and concentrated
during star formation. A natural prediction of this model is that a
similar fraction of pre-main sequence (PMS) stars of similar mass
should exhibit magnetic fields with compatible characteristics.

In order to test the fossil field hypothesis, we have performed a
sensitive spectropolarimetric survey of more than 130 field and young
cluster Herbig Ae/Be stars, the PMS progenitors of the MS A and B
stars, using the new spectropolarimeter ESPaDOnS at the Canada-France-
Hawaii Telescope, as well as its twin, Narval, at the Bernard Lyot
Telescope at Pic du Midi Observatory (France). We have discovered
strong (~kG), organised magnetic fields in approximately 5% of our
targets, and we are able to place upper limits below 100 G on the
surface magnetic fields of some undetected stars. In that talk I will
present the general results of this investigation, as well as their
implications for our understanding of the origin of magnetic fields
in intermediate-mass stars.