Four experiments were conducted to examine effects of GnRH administered at estrus on various reproductive characteristics in repeat-breeding dairy cows (eligible for third service). In Exp. 1, cows (n = 8 per group) received (i.m.) either saline or 50, 100, or 250 mug of GnRH at 12 h after onset of estrus. There was a positive linear (P < .05) effect of dose on concentrations of LH, but not of FSH, in serum at 2 h after the injection. Concentrations of progesterone (P4) were higher (P = .10) through 16 d after estrus in nonpregnant cows that returned to estrus 18 to 24 d after GnRH treatment than in nonpregnant cows given saline. Likewise, among cows with elevated concentrations of P4 through 30 d after estrus (diagnosed pregnant by increased concentrations of pregnancy-specific protein B; PSPB), P4 was higher (P = .10) in those given GnRH than in those given saline. In Exp. 2, concentrations of LH, FSH, and estradiol-17beta (E2) were similar among 12 cows during the periestrual period before they received (i.m.) 100 mug of GnRH or saline at estrus. Concentrations of P4 after estrus were increased 2 d earlier (P < .05) in GnRH- than in saline-treated cows. Among cows with elevated concentrations of P4 and PSPB through 40 d after AI, P4 was higher (P < .05) in those given GnRH than in those given saline. In Exp. 1 and 2, a greater proportion (43 vs 14%; P = .07) of cows given GnRH was pregnant 42 to 56 d after estrus than of those given saline, but calving rate was similar (27 vs 14%). In Exp. 3, average number of LH pulses per 8 h was similar between saline- and GnRH-treated cows on d 1, 3, and 8, but overall concentrations of LH were reduced (P < .05) in GnRH-treated cows. Concentrations and number of pulses of FSH were increased (P < .05) on d 8 after treatment with GnRH. Concentrations of P4 were increased earlier (P < .05) after estrus in GnRH-treated (69 +/- 12 h) than in saline-treated cows (126 +/- 12 h), with higher (P < .05) concentrations of P4 on d 4 to 8. In Exp. 4, eight cows received either 100 mug of GnRH or saline at 12 h after onset of estrus (d 0) in a cross-over design. In vitro basal production of P4 (micrograms per gram) by luteal slices from d-10 corpora lutea (CL) was similar between treatments before and after 2 h of incubation, but luteal slices from cows previously treated with GnRH produced less (P < .05) P4 in the presence of LH than did tissue from saline-treated cows. Corpora lutea had a greater (P < .05) proportion of large luteal cells (31 vs 14%) and a lesser (P < .05) proportion of small luteal cells (69 vs 86%) in GnRH- than in saline-treated cows. We concluded that administration of GnRH at estrus in repeat-breeding dairy cows increased concentrations Of P4 in serum earlier after ovulation and maintained higher concentrations of P4 for up to 40 d after treatment (during pregnancy), which seemed to be associated with higher embryonic survival until 42 to 56 d after AI. Increased P4 after GnRH was due to an increased proportion of large luteal cells in the CL and possibly due to increased concentration and pulse frequency of FSH secretion, whereas reduced in vitro production of P4 in response to LH was due to fewer small luteal cells in the CL of GnRH-treated cows.