Recently, Schmitt et al. (2010) reported an interesting experiment designed in order to test hypotheses that stress-related hormones (plasma adrenocorticotropic hormone (ACTH), cortisol, and aldosterone): (1) follow a diurnal rhythm pattern, (2) become elevated with out-of-water physical examination, (3) are not elevated during or following periods of wading human contact, and (4) plasma ACTH is a more labile and sensitive indicator of stress response-related changes than cortisol in three adult beluga whales, Delphinapterus leucas. In particular, the authors described three experiments in which the parameters were collected in a repeated-measurements design. In the first experiment, three times per day for 5 d; in the second experiment, two times per day for 5 d; and in the third experiment, five times during 24 h. The authors used Kruskal-Wallis and Mann-Whitney tests for comparison as if the experimental groups were independent.It is well known that observations may not be independent if repeated measurements are taken on the same subject. Therefore, the results of Schmitt et al. (2010) are compromised by a classical case of temporal pseudo-replication. The term pseudoreplication was coined by Hurlbert (1984), which he defined as ". . .a particular combination of experimental design (or sampling) and statistical analysis which is inappropriate for testing the hypothesis of interest." Pseudo-replication typically occurs when the number of observations or the number of data points is treated inappropriately as independent replicates. A correct statistical analysis for data from the paper of Schmitt et al. (2010) can be performed by using two-way Repeated Measures ANOVA (RMANOVA) with two within-subject variables for the first and second experiments, if normal distribution and identity matrix are satisfied. If these two assumptions are not met, Friedman's test for two-way ANOVA by ranks should be used. For the third experiment, one-way RMANOVA or Freidman's test in oneway application can be used. Choosing between two tests was described above (see