Introduction. Conditions affecting the maxillary sinus are common. This study compared the detection rates of various levels of sinus opacification using four CBCT projections. Methods. Anonymous CBCT datasets of 40 subjects were used to generate four imaging series: (1) A single ray sum simulated Waters' view [RS], (2) Bilateral multiplanar reformatted linear oblique-sagittal views [MPR], (3) Coronal 3-slice mini-series [MS] and, (4) 3D montage with selected clipping [3D]. Twelve observers subjectively rated the degree of mucosal thickening in each sinus according to a four point scale: 1=clear sinus; 2=up to 33%; 3=up to 66%; 4=greater than 66%. Results were compared with assessments made in all three orthogonal planes. Results. MS [67.5%] was superior to all other modalities in terms of correct responses (RS=52.5%, MPR=50% and 3D=46.7%). Conclusion . The use of modified projections for the detection of sinus disease is relatively inaccurate compared to examination of the entire dataset.