The provision of a truly-representative cellular sample is the major prerequisite for reliable cervical cytology. We performed a trial with a cervical-sampling implement of innovative design, the Cervex Sampler. By means of the parameters of patient and doctor acceptability, the quality of the smear, the presence of endocervical cells, and the rate of squamous-cell atypia, the Cervex Sampler was compared with a modified Ayre spatula. One hundred samplers were distributed to each of 10 doctors. The parameters were assessed before, during and after the trial period. The consensus was that the Cervex Sampler was easier to use and did not cause patients any problems. Assessment of the smears that were made with the Cervex Sampler showed a significant improvement in the quantity and evenness of the distribution of the cellular sample. During the trial, statistically-significant increases occurred in the percentage of smears that contained endocervical cells (P = 0.0001) and atypical squamous cells (P = 0.028). At the end of the trial, when the doctors again used the modified Ayre spatula, these rates reverted to the same levels as before.