Missing elements in experimental data often occur in ecological and biological sciences. In this case, it is diffi cult to carry out any data analysis and their evaluation. This paper presents one of the chemometric techniques – principal component analysis (PCA) – used to classify water quality indices on data that contain missing elements. The surface water of Czajcze Lake in Wolin National Park (northwestern Poland) was investigated. Sixteen water-quality indices were appointed in a period from April to October during 1983-2013. Conducted analysis of experimental data by EM-PCA grouped the presented water quality indices in natural clusters, including several principal components (PCs) about similar features. EM-PCA applied in the present work shows that this method can be used to analyze experimental data with missing data on considerable seasonal changes.