Navigating in the environment seems to offer the viewer a rich sensation containing a wealth of information about different objects. However, in a controlled experimental setting it is possible to demonstrate that the short-term memory is severely limited in a human observer. This rich experience is in sharp contrast with the three to four objects that can be retained in visual short-term memory. It has been known for a long time that the human visual short-term memory is limited and within the last 10-15 years a consensus on the number of objects that can be retained seems to have been reached. The focus on a limitation in terms of number of objects has probably formed the basis for a description of short-term memory as a store consisting of a limited number of slots where objects can be encoded. However, whether there is empirical evidence to support the hypothesis that the short-term memory consists of a fixed number of slots is still open for debate.<br/>This thesis presents two studies that investigate how manipulations of training and expertise (Study 1) or the general level of activation (arousal; Study 2) influence the capacity limitation in short-term memory.<br/>In Study 1 the effects of training are investigated. Since it is time consuming to achieve a level of expertise where effects conceivably would show, groups with different levels of expertise were studied in a change detection task. By contrasting stimulus material with which participants had received different levels of overt training with stimulus material that was novel for all participants, it was demonstrated that short-term memory capacity is affected by level of expertise. The results indicate that the number of objects that can be retained in visual short-term memory is influenced by the strength of the mental representations of the observer; as categories become more established fewer resources seem to be required (as reflected in a larger short-term memory capacity).<br/>Study 2 is based on an experimental investigation of repeated measures of short-term memory and investigates whether modulation of arousal affects the number of objects maintained in short-term memory. This manipulation also appears to systematically influence the number of objects encoded into short-term memory. As arousal increases the number of objects that can be maintained in short-term memory is lowered. The data suggest that this effect reflects a change in the focus of attention, which seems to narrow during heightened arousal. Two mechanisms are proposed to account for how arousal influences attentional processes; the scaling mechanism and the accentuation mechanism.<br/>Both studies suggest an embrace of the concept that objects are maintained in short-term memory. However, short-term memory should not be viewed as a static concept consisting of a fixed number of slots. The number of objects that can be maintained in short-term memory is determined partly by how well the object is known and partly by the general level of activation (arousal) in an observer.