In this paper the possible application of an electronic nose to the analysis of urine is presented. In contrast with the conventional applications of sensors and biosensors operating in liquid, the approach discussed here makes use of gas sensors performing an analysis of the headspace. The application deals with urine samples from patients affected by kidney diseases; some of the samples contained traces of blood. Results show the possibility of distinguishing the samples containing blood from the others, and a linear correlation between the first three principal components and the blood content was found. Furthermore, the electronic nose matched with a suitable neural network showed good performance in measuring the pH and the specific weight of the samples.