Recently, the initial 2D physical characterization of a small-diameter positron emission tomograph, designed specifically for scanning of small laboratory animals, was reported. The physical characteristics of the tomograph operating in 3D mode have now been measured and compared to those obtained in 2D mode. In 3D, the transaxial resolution was measured as 2.4 +/- 0.1 mm full width at half maximum (FWHM) and 6.7 +/- 1.1 mm full width at tenth maximum (FWTM) at the centre of the transaxial field of view (FOV). These values degraded to 4.5 +/- 0.3 mm (tangential) and 6.6 +/- 0.3 mm (radial) FWHM and 10.6 +/- 1.6 mm (tangential) and 11.2 +/- 2.1 mm (radial) FWTM, respectively, at a distance of 40 mm from the centre of the transaxial FOV. The axial resolution was measured as 4.6 +/- 0.1 mm FWHM and 15.0 +/- 0.5 mm FWTM at the centre of the transaxial FOV, increasing to 5.0 +/- 0.1 mm FWHM and 18.8 +/- 3.7 mm FWTM at a radial distance of 40 mm from the centre of the transaxial FOV. These resolutions are similar to those obtained for the tomograph operating in 2D mode. The sensitivity of the tomograph operating in 3D was 4.31 x 10(4) counts s-1 MBq-1 at 250-850 keV compared to 0.995 x 10(4) counts s-1 MBQ-1 in 2D at the same energy thresholds. In this energy window the noise equivalent count rate peaked at 4.1 x 10(4) counts s-1, compared to 1.03 x 10(4) counts s-1 in 2D. A scatter fraction of 30.2% at 250-850 keV was measured for a 18F line source centered in a 60 mm diameter water filled phantom in 3D, compared to 31.0% in 2D for the same scanning geometry and energy thresholds. A comparison was made between 2D and 3D kinetic analyses for a group of five anaesthetized rats scanned using [11C] SCH 23390, a marker of dopamine D1 receptors. The integrity of the results was maintained between 2D and 3D data sets, though in 3D there was a significant reduction in the standard error of the fitted parameter. The results demonstrate that, with regard to sensitivity, there are significant gains in the physical performance of this tomograph when operating in 3D compared to 2D mode and that the quantification of PET studies of small animals using the 3D data reflects this.