A retrospective audit of 1079 2-week referrals between 1 January 2004 and 31 December 2006 was undertaken. The aims of this audit were to assess compliance of referrals with Department of Health (DoH) guidelines; the effectiveness of the 2-week referral route in detecting head and neck cancers, and to determine whether this route identified more early stage cancers. Of 1079 2-week referrals, 71.5% conformed to DoH criteria. DoH guidelines were found to have a high sensitivity of 83.9% (75.5-89.7%, 95% CI) for head and neck cancer, but a low positive predictive value of 12.8% (10.5-15.3%) and a specificity of 30.0% (27.2-33.1%). Only 10.9% of 2-week referrals were diagnosed with a head and neck cancer. The cancer detection rate was higher amongst referrals that conformed to DoH guidelines (12.8%) compared to those that did not 6.2%. This was statistically significant (Chi square, p<0.01). The guidelines had a positive likelihood ratio of 1.20 (1.1-1.3), suggesting that there is a minimal increase in the likelihood of head and neck cancer when DoH guidelines are correctly applied. The diagnostic odds ratio (DOR) of the DoH referral criteria is 2.21. Most head and neck cancers were diagnosed via routine referral routes, 2-week referrals contributing to only 21.4% of all head and neck cancers diagnosed during the study period. The 2-week referral route did not identify more early stage cancers.