Addition of CanoDraw for Windows represents the major enhancement of Canoco (TM) statistical package in its version 4.5. CanoDraw supports thoroughly all the visual aspects of multivariate data analysis with linear and unimodal ordination methods. Analysis of multivariate data often involves not only the response variables, but also the explanatory variables. Canoco recognizes at least three kinds of explanatory variables, depending on the role assigned to them in particular analysis: environmental variables, covariables, and supplementary variables. Ecologists often ask several different questions about the same datasets, and the individual questions are reflected in separate Canoco projects. CanoDraw supports this project-oriented approach to data analysis, so it is also centered around projects, stored in files using a CDW extension in their names. These projects correspond one-to-one with Canoco projects. Unlike the Canoco project files (which have the CON extension in their names), CanoDraw project files contain not only the information about analysis properties, but also the original analyzed data, ordination results produced by Canoco, as well as any additional project properties defined by the user in the CanoDraw program (groups of objects, objects classifications, etc.)