Semiquantitative immunoassay technology, in the form of rapid test strips, offers a less time-consuming and less costly alternative to other methods of verifying self-reported smoking status, such as gas chromatography-nitrogen phosphorus detection (GC). Unfortunately, information on the validity and reliability of some test strips in urine and saliva samples is not always available. This paper describes the diagnostic accuracy of one type of test strip currently available (NicAlert cotinine test strips; NCTS). GC was used as the reference standard and saliva as the sample medium. The study involved 86 people (41 smokers and 45 nonsmokers) aged 18 years or over, who were able to understand written English and provide written consent. Pregnant women, women with infants less than 6 weeks old, and people who had eaten 30 min prior to sample collection were excluded. Two saliva samples were collected simultaneously from each participant, with one sample tested using NCTS and the other by GC analysis. People with at least 10 ng/ml cotinine (in both tests) in their saliva were considered smokers. NCTS were found to have a specificity of 95% (95% CI 89%-100%), a sensitivity of 93% (95% CI 85%-100%), a positive predictive value of 95% (95% CI 89%-100%), and a negative predictive value of 93% (95% CI 86%-100%). The use of NCTS is a valid and reliable method, compared with GC, to test saliva samples for verification of smoking status.