A new test of olfactory function, the combined olfactory test, has been designed to assess odours easily recognizable by the test population. The test consists of an odour recognition test of nine odours, where an odour in a bottle is chosen from a list of four possible odours in a forced choice manner. This is followed by a threshold test using a series of three-fold dilutions of 1-butanol. The mean of the two scores is the combined olfactory score. The test was subjected to a validation study. It was performed on 133 participants with a normal sense of smell and a normal rhinological examination and on 94 participants who said that they did not have a sense of smell. There was a highly significant difference between the combined olfactory score in the normal and 'anosmic' groups (P < 0.001). This significant difference was the same between the two groups for the threshold and odour recognition arms of the test. There was a highly significant difference (P < 0.001) between the two subgroups of 'completely anosmic' and 'almost anosmic' participants, indicating that the test could grade the degree of olfactory dysfunction.