In 1972, Papi and his colleagues reported that anosmic pigeons were severely impaired in orientation and homing performance. This observation was followed up in a series of experiments involving numerous elaborate experimental manipulations. On the basis of their results, the hypothesis of olfactory navigation was proposed. Attempts to replicate these findings at other lofts produced widely differing effects, which suggested a highly variable role of olfaction. However, meteorological data, as well as certain other aspects of the findings, throw doubt on the role of odours as navigational cues. (1) Odours of the required characteristics and distribution do not seem to exist. (2) Some effects of 'olfactory' manipulations do not seem to depend on the availability of odours. (3) Olfactory treatments proved mostly effective, but often the effect was not as predicted. In view of these findings, explanations other than olfactory orientation cannot be excluded; accepting olfactory input as navigational information seems premature. Some of the findings are in agreement with the assumption that olfactory manipulations impair the birds' general processing and integration of information in some unknown way.