Objective To determine if using the endocervical brush after curetting the endocervix will increase the yield of endocervical tissue retrieved for an endocervical curettage (ECC) specimen. Methods Between March 1, 1995, and June 30, 1996, we recruited for participation patients with abnormal Papanicolaou smears referred for colposcopy. Exclusion criteria were pregnancy, previous hysterectomy, and a history of diethylstilbestrol exposure. Colposcopy and biopsies were performed by residents under direct supervision of the attending staff. Endocervical curettages were performed using the Kevorkian endocervical curette. The subjects were then assigned randomly to one of two ways of collecting the ECC tissue: with either a curette or an endocervical brush. Specimens were reviewed by pathologists, who were blinded to the method of ECC collection. Results During the study period, 124 patients agreed to participate; 62 were assigned to the control group, and 62 to the study group. Six subjects had missing data, leaving 118 patients available for analysis. In the control group, six of the 58 ECC samples obtained contained insufficient endocervical tissue for pathologic diagnosis. None of the 60 samples from the endocervical brush group was insufficient. The difference between the two groups was statistically significant ( P = .01). Conclusion The addition of the endocervical brush to endocervical tissue sampling at colposcopy in the study decreased the number of insufficient samples. The endocervical brush method of collection of an ECC specimen from the canal after the Kevorkian curette is used is a valuable addition to this diagnostic tool. We recommend its use in obtaining an ECC specimen.