Atrial myocytes differ from ventricular myocytes in ultrastructure and calcium handling proteins and consequently show subtle differences in calcium signalling. However there is little quantitative information characterising atrial calcium signalling. This study compares atrial and ventricular calcium transport focussing on the removal of calcium from the cytosol by sarcoplasmic Ca-ATPase (SERCA). All procedures were approved by the University of Bristol ethics committee and conformed to the Animals (Scientific Procedures) Act, 1986. Rabbit atrial and ventricular myocytes were isolated by enzymatic digestion. Isolated cells were loaded with fluo-4, electrically stimulated at 1 Hz and line-scan images acquired using a Zeiss Pascal LSM5 laser-scanning confocal microscope. Rates of decay were obtained by fitting the recovery phase of calcium transients with a single exponential decay. Electrical stimulation of atrial and ventricular cells resulted in calcium transients, which were homogenous across the cell in ventricular myocytes, but were consistent with preferential coupling between calcium entry and release at the edge in atrial cells. However, there was no significant difference in the rate of recovery (ktwitch) of the spatially averaged calcium transients in atrial (1.8 ± 1.1 s-1, n=18) and ventricular myocytes (1.5 ± 1.1 s-1, n=11, P=0.21 un-paired t-test). Application of 10 mM caffeine, to release calcium from the sarcoplasmic reticulum and prevent its subsequent re-uptake, significantly slowed the rate of recovery (kcaff) in both cell-types. However, whereas kcaff was 50.0 ± 5.7% (n=11) of ktwitch in ventricular cells, kcaff was 14.4 ± 1.2% (n=19) of ktwitch in atrial myocytes, indicating that SERCA plays a greater role in the removal of calcium in atrial cells than in ventricular cells. When 10 mM Ni was applied in the presence of caffeine to block Na/Ca exchanger, further slowing of the rate of recovery (kNi) was observed in both cell types, although to a much larger extent in atrial cells. That kNi was faster in ventricular cells suggests that other slow pathways contribute more to calcium extrusion in these cells. To confirm the greater role of SERCA in atrial cells, the SERCA inhibitor thapsigargin (TG) was applied during electrically-stimulated calcium transients. Rate constants in the presence of 5 µM TG were significantly reduced in both atrial and ventricular cells, representing 16.8 ± 3.1% of ktwitch (n=6) in atrial compared with 49.7 ± 7.2% (n=8) in ventricular cells, consistent with the caffeine data. In summary, the data presented here indicate a greater contribution of SERCA to calcium extrusion in atrial cells compared to ventricular cells and highlights important differences in atrial and ventricular calcium signalling.