Objective: Handheld computers and mobile phones allow assessments of a variety of aspects of behaviour, experience and performance in an everyday life setting. We have evaluated the use of a mobile (cell) phone to collect data on alcohol consumption, subjective state, and psychomotor performance and to assess the relationships between these variables. Methods: Thirty eight healthy volunteers (20 male) aged 18-54 years (mean 22.8) took part. The study program ran on any phone supporting downloadable Java® applications. Text (SMS) messages were sent twice a day for 14 days according to a randomised schedule, and volunteers completed their entries as soon as possible after receiving the text. They recorded alcohol consumption and mood, and completed performance tasks. Results: Compliance was good, with responses being made to over 80% of text messages. Subjective drowsiness (% of visual analogue scale length) ranged from 35.7-42.9 between 09:00 and 19:00 and from 51.5-57.7 between 21:00 and 01:00. Mean ratings of drunkenness assessed between 21:00 and 01:00 were 5.5% (S.D. 12.2) where no alcohol consumption was recorded and 53.1% (S.D. 21.0) where five or more drinks had been consumed in the last six hours (ANOVA p<0.001). Mean errors on a memory scanning task were 6.6 (S.D. 4.7) with no alcohol and 10.5 (S.D. 6.8) after 6 or more drinks (ANOVA p<0.01) Conclusion: Both objective and subjective measures showed the expected relationship with reported alcohol consumption. The widespread use of mobile phones means that this type of study can be carried out economically and with reasonable training requirements. Mobile phones provide a very effective method of collecting data in an unsupervised, naturalistic setting.