Air leakage through the building envelope is of great importance for the energy use of a building. However, from an indoor air quality standpoint, the size of interior leaks in e.g. multifamily buildings could be important as e.g. a source of pollution. Using the standardised Fan Pressurization test method, it is not possible to separate interior leaks from leaks in the building envelope. One way to separate these leaks is to simultaneously depressurise (or pressurise) adjacent apartments to the same pressure and thereby eliminating interior leakage. The difference from the standardised measurement then represents the interior leakage. By adding one adjacent apartment at a time, leakage curves for the different building components could be established. Results from measurements in three apartments show interior leaks to be between 12 and 36 % of total leakage at 50 Pa. Because the apartments are airtight to begin with, 0.45 to 0.90 airchanges per hour at 50 Pa, the interior leakages in absolute numbers are very small and sensitive to inaccuracies in the measurements. Only a few component leakages showed a meaningful curvefit. Vertical leaks were found to be bigger than horizontal, probably owing to more penetrations in floor structures for building services. Pressure differences between apartments in a range of 5-15 Pa have been measured as an effect of operating the mechanical kitchen ventilation. At a 10 Pa pressure difference, air flow between apartments of 4-15 m³/h could be seen from the measured leakage curves. If these comparatively small air flows during limited periods of time cause any problems with odours, pollutants etc. should depend on the conditions in the adjacent apartments. Glazed courtyards are at the moment a popular feature in new and remodeled Scandinavian buildings. Very little data is available on airtightness and air infiltration regarding these courtyards. For design calculations only rough estimates are used. Air leakage of the glazed courtyard in the Suncourt building was found to be greater than airchanges per hour at 50 Pa or about 15m³/hm² leaking surface area. The average design air infiltration rate during operating conditions of 0.2 airchanges per hour might therefore be exceeded.