The principle of aggregation states that the sum of a set of multiple measurements is a more stable and representative estimator than any single measurement. This greater representation occurs because there is inevitably some error associated with measurement. By combining numerous exemplars, such errors of measurement are averaged out, leaving a clearer view ofunderlying relationships. The present study explored the effect of score aggregation over various time periods on correlations among a number of reliable measures frequently used in open-field testing. Twenty-six male rats were given four open-field tests (4 min in duration) at 48-h intervals. Ambulation, rearing, and defecation responses were measured on a minute-by-minute basis in the open-field tests. Correlation matrices were calculated among the three measures for unaggregated scores (Lrnin totals) and for scores aggregated over daily tests, and mean correlation coefficients were computed for all three pairwise comparisons of the three response variables. These mean correlations were then compared to those obtained when the open-field measures were aggregated over all 4 test days. The results showed that aggregation produced substantial increases in correlation-coefficient magnitude. The correlation between ambulation and rearing increased from a mean of .39 to a value of .81. Similar increases were observed when defecation scores were correlated with ambulation (-.17 to -.59) and rearing (-.16 to -.49). Thus aggregation is an important factor to be considered in the design of psychobiologicalcorrelational studies.In recent years, the principle of aggregation has received renewed interest in the areas of personality (Epstein, 1979(Epstein, , 1980 and behavioral development (Rushton, Brainerd, & Pressley, 1983). This principleis based upon the notion that the sum of a set of multiple measurements is a more stable and unbiased estimator than any single measurement from the set. Becauseany measurement has an error component associated with it (cf. Gulliksen, 1950), the combining of several measurements tends to average out these error components, providing a better estimate of the true value of the parameter in the population of interest. Perhaps the best-known example ofthis principle is the common rule that the larger the sample of a set of measurements, the more representative is the sample mean of the population mean.In psychobiological research, as in personality or developmental research, the principleof aggregationcan be used to obtain better estimates of the true values of various behavioral, physiological, or neurochemical parameters. A well-known rule in educational and personality testing is that the reliability of an instrument increases as the number oftest items increases (e.g., Gulliksen, 1950;Lord & Novick, 1968). Another example This study was supported by a grant from the Natural Sciences and Engineering Research Council ofCanada (UOI51) tothe first author. We thank J. P. Rushton for helpful discussion. Please send reprint requests to:...