Continuous observations made during the driving of a tunnel have always been taken for granted, just like keeping one's eyes on the road when driving a car. Therefore, the term 'observational method' often produces a reaction of surprise among tunnelling engineers in practice. Even if it is explained that 'observation' in the narrow sense of the word really means 'monitoring', doubt and misunderstanding still remain. In the first part of the paper the term 'observational method' as introduced by Peck and redefined by Eurocode is examined critically. The problems encountered in tunnelling are usually so diverse and complex that monitoring can be usefully applied in many different ways. In the second part of the paper a case study is described in which Peck's ideas are followed.