At the end of the 1970s otter surveys were developed to study the distribution of the Eurasian otter Lutra lutra. The method was quickly adopted and expanded to include other species of otter. Different methods to estimate otter density were also used. There is some uncertainty, however, as to the significance and precision of the results. Otter surveys, radio-tracking studies and otter censuses (tracks and visual) were conducted in the same areas simultaneously. Comparison of the home ranges of otters tagged with transmitters with the results from otter surveys carried out in the same areas showed very similar otter distribution. Some 71% of those sites examined in which only one otter lived were positive, and 97% of those sites with two or more otters. Differences were found between different watercourses. In small or average sites with waterways, one single otter was capable of marking many kilometres so that it could be detected in most of the conventional 600 m sites (here 71% of sites). The data obtained by radio-tracking showed that almost all otters can be seen by experienced observers, confirming the validity of visual censuses. A high correlation was found between the number of otters seen during visual censuses and the number of otters detected by means of the length of their footprints. Using new and old tracks (2-days-old or more) the number of otters was overestimated.