Genetic factors play an important role in determining bone mineral density (BMD) in later life, with the genetic influence mediated through effects on both peak mass and on age- and menopause-related bone loss. At menopause there is an increase in the production and activity of various cytokines and growth factors within the bone microenvironment. The activity of interleukin-1 (IL-1), a powerful stimulant of osteoclastic bone resorption, is increased in estrogen-deficient states with increased production of IL-1 and inhibition of the IL-1 receptor antagonist (IL-1ra). Treatment with IL-1ra blocks the bone loss associated with ovariectomy in animals and the IL-1 receptor antagonist gene (IL-1RN) is therefore a potential candidate gene for the regulation of postmenopausal bone loss. We examined the relationship between annual rates of change in BMD over 5 years and an 86 bp variable number tandem-repeat polymorphism of the IL-1RN gene in 108 early postmenopausal women. All women were within 5 years of a natural menopause at the study's onset, healthy, and not on hormone replacement therapy or other medication known to affect bone metabolism. BMD was measured annually over the 5 year study period at the lumbar spine and femoral neck using dual-energy X-ray absorptiometry. Three alleles were identified (A1 = 4 repeats, A2 = 2 repeats, A3 = 5 repeats), with five genotypes observed: A1A1 (41.7%), A1A2 (45.4%), A2A2 (6.5%), A1A3 (2.8%), and A2A3 (3.7%). For analysis, alleles were collapsed into a biallelic system grouping the A1 and A3 alleles. There was no significant relationship between the IL-1RN genotypes and baseline bone mass at either the spine or hip. IL-1RN genotype was significantly associated with annual rates of change in spinal bone mass (p < 0.05), and this finding remained significant after adjustment for age, weight, and baseline BMD. Carriage of at least one copy of the A2 allele was associated with reduced bone loss at the spine (mean change in BMD +/- SD: -0.81 +/- 1.46%/year) when compared with noncarriage of the A2 allele (mean change -1.38 +/- 1.48%/year), p = 0.05. We therefore conclude that allelic variation at the IL-1RN locus is associated with differential rates of early postmenopausal bone loss at the spine. Further research will be required to clarify the mechanisms underlying these findings and to determine whether this association translates into a significant long-term effect on BMD and fracture in later life.