The standard method for detecting meningococcal carriage is culture of throat swabs on selective media, but the levels of carriage determined depend heavily on the skills of the individuals taking the swab and interpreting the cultures. This study aimed to determine the most sensitive detection method for meningococcal carriage. Throat swabs and saline mouth gargles, obtained from 89 university students, were processed in parallel by conventional culture and TaqMan ctrA PCR. Carriage of meningococci, as detected by the combined methods, was 20%. The sensitivities of throat swab culture, throat swab PCR, gargle culture, and gargle PCR were 72, 56, 56, and 50%, respectively, and the probabilities that these techniques would correctly identify the absence of carriage (negative predictive value [NPV]) were 93.4, 89.9, 89.9, and 88.8%. Culturing both throat swabs and gargles increased the NPV to 98.6%. The further addition of throat swab PCR increased this to 100%. Testing gargles by both culture and PCR was as sensitive as testing throat swabs by both methods, suggesting that gargles may be a suitable alternative for large-scale screening studies when throat swabs are difficult to obtain, although they required more lengthy laboratory processing. PCR was a useful adjunct to culture for detecting nasopharyngeal carriage, but it failed to detect some nongroupable strains. For maximum sensitivity, a combination of techniques was required. This study indicates the confidence with which health care professionals involved in meningococcal screening can regard laboratory results.