It has been traditionally accepted that antipsychotics have a 'delayed onset' of action with various authors claiming 2–3 weeks of delay. Data from the previous few years has contradicted this notion.

The talk will review this data, identify its theoretical implications and implications for practise. In an initial meta-analysis of over 8,000 patient data it was shown that antipsychotic action is not delayed at all but evident in teh first week, this change is not non-specific sedation but a specific anti-'psychotic' effect, and is seen with all evaluated antipsychotics. In more recent studies we have been able to show that this effect is even discernable in the first 24 hours, is related to dopmaine D2 occupancy achieved early on, and early non-response is a very strong predictor of subsequent non-response.

This early onset is not only an isolated improvement in psychosis-but, can also be seen on quality of life and social functionign indicators. These data raise interesting questions regarding the mechanism of improvement: they question the standard 'depolarisation block' hypothesis of antispychotic action-and suggest a simpler dopamine occupancy hypothesis instead; since change occurs early it is possible to map different trajectories and our most recent findigns show that drug-treated patients show a distinct trajectory from those observed in placebo-treated patients. Finally, these findings have implications for the design of clinical trials and clinical practise. In clinical trials it is possible to use early-response to devise adpative or enriched designs that are more clinically relevant - example of a recent such application will be discussed and at a clinical level it may be possible to reconsider the standard notion that an antipsychotic must be tried for 6 weeks before considering an alterative option.