The pharmacokinetics of recombinant human relaxin (rhRlx) after intravenous (iv) bolus administration and the absorption of rhRlx after intracervical or intravaginal administration were determined in nonpregnant women. The study was conducted in two parts. In part I, 25 women received 0.01 mg/kg rhRlx iv. After a minimum 7-day washout period, these women were dosed intracervically (n = 10) or intravaginally (n = 15) with 0.75 or 1.5 mg rhRlx, respectively, in 3% methylcellulose gel. Part II was a double-blind, randomized, three-way crossover study in 26 women. At 1-month intervals, each woman received one of three intravaginal treatments consisting of 0 (placebo), 1, or 6 mg rhRlx in 3% methylcellulose gel. The serum concentrations of relaxin following iv administration were described as the sum of three exponentials. The mean (+/- SD) initial, intermediate, and terminal half-lives were 0.09 +/- 0.04, 0.72 +/- 0.11, and 4.6 +/- 1.2 hr, respectively. Most of the area under the curve was associated with the intermediate half-life. The weight-normalized clearance was 170 +/- 50 mL/hr/kg. The observed peak concentration was 98 +/- 29 ng/mL, and the weight-normalized initial volume of distribution was 78 +/- 40 mL/kg, which is approximately equivalent to the serum volume. If central compartment elimination was assumed, the volume of distribution at steady state (Vss/W) was 280 +/- 100 mL/kg, which is approximately equivalent to extracellular fluid volume. Vss/W could be as large as 1300 +/- 400 mL/kg without this assumption.(ABSTRACT TRUNCATED AT 250 WORDS)