In May 1996 a conference entitled 'Climate and Ozone at the Dawn of th
e Third Millenium' was held in Brussels. Paul Crutzen (Nobel Prize for
Chemistry, 1995), Bert Bolin (Blue Planet Prize, 1995), Willy Dansgaa
rd and Nick Shackleton (co-recipients of the Crafoord Prize, 1995) led
the debate both to celebrate the extraordinary recent achievements of
European environmental research as well as planning how to continue t
his success in the future. The conference covered the whole range of E
uropean environmental science (Berger et al., 1996). What we have summ
arised here is the debate on the future direction of palaeoclimatic re
search. It was suggested that priorities for palaeoclimatic research s
hould be set in three ways: (1) what data do climate models require fr
om the palaeoclimatologists; (2) as a feedback, what models do the pal
aeoclimatologists require; and (3) there is still a need for pure or n
on-directed research as this has in the past provided significant insi
ghts into the way the climate systems work? The delegates also agreed
that there was a need for palaeoclimatic data on three different time
scales and resolutions: (1) data for the last 200-500 years with a res
olution of at least a month Iso that climate models can be tested and
natural rapid climate changes such as the El Nino-Southern oscillation
can be investigated; (2) data for at least the last 2000 years if not
the whole of the Holocene at yearly resolution to investigate natural
climate variability; and (3) longer term palaeoclimatic studies to in
crease our general knowledge of climate change. (C) 1997 Elsevier Scie
nce Ltd.