Abstract Our purpose was to examine whether present sound insulation guidelines of private office rooms provide optimal cognitive performance and acoustic satisfaction. 32 voluntary students participated in our laboratory experiment. The experiment simulated two adjacent office rooms. Speaker was in room 1 and listener in room 2. Both sound insulation between the rooms and the background noise level (sound masking) in room 2 were investigated at two levels so that the speech intelligibility between the rooms was significantly changed. Condition A corresponded with the present Finnish sound insulation guidelines according to which the weighted sound reduction index should be at least 35 dB R ' w and background noise level from building services should not exceed 33 dB L Aeq . Conditions B and D had a 10 dB greater sound insulation than condition A. Conditions C and D had 9 dB larger background noise level in room 2 than condition A. Cognitive performance was the worst in condition A. Acoustic satisfaction and several other subjective measures showed that conditions B-D were better than condition A. The present sound insulation guidelines need to be reconsidered to provide better acoustic satisfaction and less distraction for employees in private office rooms.