Relationship among energy balance (EB), first ovulation, circulating metabolites and opioid involvement in peripheral LH concentrations were assessed in 40 multiparous Holstein cows assigned randomly to an experiment with a 2 x 2 factorial arrangement of treatments. Cows were either milked normally (2x/d) and expected to be in negative EB (LAC) or were milked for only 1 d postpartum (PP) to ensure a rapid return to positive EB (NONLAC). Blood samples were collected from NONLAC cows every 12 min for 16 h on d 7, 10, and 14 PP and from LAC cows on d 7, 14, 21 and 28 PP. Within each group, half the cows received naloxone (50 mg/h) and half received saline during h 9 to 16 of each frequent sampling series. The mean number of LH pulses/8 h for NONLAC cows was 4.3, 4.7 and 5.3 at 7, 10 and 14 d PP and for LAC cows 4.3, 5.7, 6.4 and 7.0 at 7, 14, 21 and 28 d PP, respectively. The LH pulse frequency was not different between NONLAC and LAC cows at 7 and 14 d PP, yet NONLAC cows had fewer days to first ovulation (P less than .01). Naloxone did not affect any parameters of LH secretion in either group. Following parturition, NONLAC cows averaged 4.0 d to negative EB nadir and 14.3 d to first ovulation. The LAC cows averaged 13.6 d to negative EB nadir and 27.0 d to first ovulation. Days PP to first ovulation were highly correlated (r = .85) with days PP to negative EB nadir.(ABSTRACT TRUNCATED AT 250 WORDS)