In emission tomography, the spread of regional tracer uptake to surrounding areas caused by limited spatial resolution of the tomograph must be taken into account when quantitating activity concentrations in vivo. Assuming linearity and stationarity, the relationship between imaged activity concentration and true activity concentration is only dependent on the geometric relationship between the limited spatial resolution of the tomograph in all three dimensions and the three-dimensional size and shape of the object. In particular it is independent of the type of object studied. This concept is characterized by the term "recovery coefficient". Recovery effects can be corrected for by recovery coefficients determined in a calibration measurement for lesions of simple geometrical shape. This method works on anatomical structures that can be approximated to simple geometrical objects. The aim of this study was to investigate whether recovery correction of appropriate structures is feasible in a clinical setting. Measurements were done on a positron emission tomography (PET) scanner in the 2D and 3D acquisition mode and on an analogue and digital single-photon emission tomography (SPET) system using commercially available software for image reconstruction and correction of absorption and scatter effects. The results of hot spot and cold spot phantom measurements were compared to validate the assumed conditions of linearity and stationarity. It can be concluded that a recovery correction is feasible for PET scanners down to lesions measuring about 1.5xFWHM in size, whereas with simple correction schemes, which are widely available, an object-independent recovery correction for SPET cannot be performed. This result can be attributed to imperfections in the commercially available methods for attenuation and scatter correction in SPET, which are only approximate.