The present study was designed to compare quantitative variance estimates in the profile of the luteinizing hormone (LH) surge within individual rats over successive proestrous (PE) days (WITHIN) with the variability between rats (BETWEEN). Sprague-Dawley female rats were implanted under ether anesthesia with indwelling intracardiac cannulas. On successive PE afternoons of normal 4-day cycles, hourly blood samples (0.25 ml) were collected via the cannula from 1400-2000 h for radioimmunoassay of plasma LH. Three characteristics which reflected the profile of the LH surge were examined: the time of onset of LH release, the time of peak LH release, and the magnitude of peak LH release. Twenty-one animals yielded LH surge data on a total of 42 PE days with a mean (+/- SD) time of onset = 1534 h +/- 66 min, time of peak = 1730 h +/- 75 min, and magnitude of peak = 1176 +/- 441 ng NIAMDD-Rat LH-RP-1/ml plasma. Variance estimates BETWEEN and WITHIN animals were determined by analysis of variance and the method of Vaughan and Corballis (1969) for calculating percent of total variance. Differences BETWEEN in time of onset of LH release approached significance (P = 0.05-0.10) and contributed 50.5% of the total variance compared to a negative value for differences WITHIN. Differences BETWEEN in time of peak LH release were significant (P less than 0.05) and contributed 51.5% of total variance compared to a negative value for differences WITHIN. In contrast, for the magnitude of peak LH release, neither differences BETWEEN nor WITHIN contributed substantially to total variance (both negative values), with the major contribution from the residual term.(ABSTRACT TRUNCATED AT 250 WORDS)