The effectiveness of three deep breathing techniques was evaluated in 98 male patients after coronary artery bypass graft surgery in a randomized trial. The techniques examined were deep breathing with a blow bottle-device, an inspiratory resistance-positive expiratory pressure mask (IR-PEP) and performed with no mechanical device. Pulmonary function and roentgenological changes were evaluated. Four days post-operatively there were significantly decreased vital capacity, inspiratory capacity, forced expiratory volume in 1 second, functional residual capacity, total lung capacity and single-breath carbon monoxide diffusing capacity in all three groups (p < 0.0001). No major differences between the treatment groups were found, but the impairment in pulmonary function tended to be less marked using the blow bottle technique. The Blow bottle group had significantly less reduction in total lung capacity (p = 0.01) compared to the Deep breathing group, while the IR-PEP group did not significantly differ from the other two groups.