In quantifying the effects of mild sedatives both physiological and subjective aspects of sleep must be taken into account. A questionnaire analysis on a mild sedative (400 mg of an aqueous extract of Valeriana officinalis L.) showed that by subjective criteria it is sedative (i.e. it significantly decreased perceived sleep latencies and night awakenings, and improved sleep quality). In an EEG study on the same preparation the pattern of results tended to confirm the subjective evaluation (i.e. shorter mean sleep latency, increased mean latency to first awakening) but the changes did not reach statistical significance. The discussion critically examines some of the approaches used to test putative mild sedatives and suggests a rational approach to analysing their effects.