During the period 20-8 ka BP, movements of the polar front in the North Atlantic Ocean between the latitudes of Iceland and the Iberian peninsula greatly affected the climate of western Europe. During the Lateglacial, sea-surface temperature changes were particularly marked in the Bay of Biscay. Such migrations of the polar front, which have been shown to be time-transgressive, have been used to explain Lateglacial climatic events in northwestern Europe. A comparative study of Lateglacial and early Holocene records from lacustrine sites in northern and northwestern Spain and the Pyrenees confirms that the Lateglacial climatic amelioration was time-transgressive along the seaboard of western Europe, beginning 500-1000 years earlier in northwestern Spain than in the British Isles. This time-lag is further exaggerated in the vegetational response by migrational lags and edaphic factors. There are marked differences in the nature and chronology of Lateglacial plant successions, not only between southwest and northwest Europe, but particularly between sites in northwestern Spain, the coastal lowlands of the Pays Basque and the Pyrenees. Sites in northwestern Spain, including that of Sanabria Marsh, here published in detail for the first time, show the moderating climatic influence of the Atlantic Ocean throughout the Lateglacial. There, the climatic amelioration began early, perhaps before 14 ka BP. Deciduous oak forest had already begun to develop during Lateglacial times; this observation suggests that the perglacial refugia for these trees lay close to the maritime Atlantic coasts of Spain and Portugal, and not in the Pyrenees as some authors have proposed. After the onset of the Lateglacial climatic amelioration, pine and birch forest became widespread in the Pyrenees but oaks were very sparse or absent. Oak forest only developed there after 10 ka BP in the early Holocene. The Younger Dryas episode of cooling can be detected, but only by a small expansion of herbaceous plant communities in some areas and with almost no lowering of the treeline. In contrast, Lateglacial conditions in the Pays Basque appear to have been cold and bleak. Even birch and pine forest was poorly developed and may have disappeared with the onset of the Younger Dryas cooling. Acid heathland with Empetrum and ericaceous plants then developed, to be replaced by oak-hazel forest in the early Holocene. Here, clearly, the influence of cold polar water conditions in the Bay of Biscay was very strong. Pollen diagrams from marine cores in the Bay of Biscay are also reviewed, but low sedimentation rates, bioturbation and differential transport and preservation of pollen make comparison with continental pollen diagrams difficult and correlation only possible in broad terms. Accurate vegetational interpretations are impossible. Palynologists working on archaeological cave and rock shelter sequences in southwest France and northern Spain have claimed to recognize, between 32 and 14 ka BP, a series of interstadial intervals with expansions of temperate trees. Careful consideration of pollen diagrams covering the purported Laugerie and Lascaux interstadials, said to occur between 16 and 20 ka BP (conventionally the maximum period of glacial advance of the last glacial stage), suggests that temperate pollen has percolated down through overlying deposits and been preserved in certain sedimentologically favourable beds. Although widely accepted by archaeologists, these interstadials appear to have no reality and must be rejected. There is no trace of them in the long lacustrine records of Les Echets (Beaulieu & Reille 1984) and Grande Pile (Woillard 1975, 1978). There is thus no good palynological record for 30-16 ka BP from south-west Europe, other than the long pollen sequence from Padul in southern Spain (Pons & Reille 1986).