The detection of truthful statements has been investigated over the years using physiological, nonverbal and verbal cues. Researchers have been trying to establish what method of detection is most reliable, with contradictory evidence being found both for and against each of these methods of truth detection. The Aberdeen Report Judgment Scales (ARJS) enable the verbal content of a statement to be assessed using specific criteria thought to be indicative of a self-experienced account. Two experiments were carried out to investigate the ability of the ARJS criteria. Experiment la consisted of 48 accounts of either a pleasant or an unpleasant evening dinner. Half of the accounts were true while the other half were false. Those using the criteria, along with outcome feedback, were expected to rate self-experienced accounts as more credible than transformed accounts in comparison to those participants who did not receive guidance. No significant differences in credibility ratings were found between conditions or in relation to truth status. Experiment lb consisted of 60 accounts of an overnight military officers training exercise. Again, half the accounts were self-experienced while the remaining accounts were invented. Within these two conditions, half of the account tellers had been given a short preparation time while the other half had been given a long preparation time before presenting their accounts. In addition, the effects of feedback on credibility ratings were investigated. A significant difference was found for credibility ratings of self-experienced and invented accounts in the short preparation condition only. No differences were found between groups. These findings, along with the implications they have on use of the ARJS scales are discussed and modifications to the present research are suggested.