In behavioural and evolutionary ecology, there are often large phenotypic differences between individuals in, for example, body size or large variation in abiotic conditions such as temperature, between measurements. This often inevitable source of variation may mask any effect of experimental treatment as it can have a large impact on the dependent variable of interest. In such cases, conventional statistical comparisons may have much lower power than desired. The inclusion of covariates in statistical analyses has proven a powerful method to control for such nonrandom differences between individual data points that cannot be controlled experimentally (Huitema 1980). To make correct conclusions, it is important to understand the basic assumptions underlying such a covariate analysis. In this paper I argue that this has evidently not been completely acknowledged in the scientific community. Sophisticated models relating responses to both one or more continuous covariates and one or more factors can be problematic. Factor is here used in the meaning of a categorical independent variable and its value divides individuals into discrete groups or categories, for instance experimental treatments. In the following, I use a simple one-factor ANCOVA design as an example, but the same general problem outlined here applies to all linear models with one or more covariates, including generalized linear models (GLIM) such as logistic regressions and even survival analysis.The basic design of a one-factor fixed effect ANCOVA can be written as:where Y ij denotes the values for the dependent response variable of the jth subject in the ith category of the factor, m is the mean intercept (the average value of the response parameter when the value of the covariate equals zero), a i is the response to the ith category of the factor, X ij the value for the covariate of the jth subject in the ith category of the factor, X is the mean value of the covariate for all individuals, b is the overall pooled regression coefficient (slope) within groups and e the normally distributed error variance (cf. Huitema 1980). When performing an ANCOVA, we thus assume equivalent slopes among treatment groups (b). The test of homogeneity among slopes is therefore a key prerequisite to proceed to the ANCOVA itself. The easiest way to test this assumption is to include the interaction term between the covariate and the factor in the model. If the interaction term is nonsignificant, we can conclude that the slopes are homogeneous and then proceed to test whether the response differs between groups. This is formally done by testing for differences between treatment groups in the Y intercept for the regressions of the covariate on the response variable Y. This test is carried out by re-running the model, excluding the interaction term. Differences in the Y intercept (i.e. differences between groups when the value for the covariate equals zero) are generally of minor importance. However, since we assume homogeneity of slopes this difference can be inferred o...