Family studies have demonstrated striking differences between individuals in their ability to produce IL-10 following lipopolysaccharide (LPS) stimulation of whole blood cultures in vitro, suggesting that differences in IL-10 production involve a considerable hereditary component. The first aim of this study was to analyse the possible effect of IL-10 genotypes and haplotypes on IL-10 plasma levels in a healthy Finnish population. As previous reports have demonstrated that endogenously produced IL-1 induces LPS-stimulated IL-10 production and that IL-10 inhibits synthesis of IL-1 in human monocytes, it is apparent that these two cytokines form an autoregulatory feedback loop. Secondly, we were interested whether any relationship could be found between IL-10 and IL-1beta in vivo. To examine this, the influence of IL-1alpha -889, IL-1beta -511 and IL-1Ra VNTR genotypes and IL-10 genotypes/haplotypes (ACC, GCC and ATA) on IL-10 plasma levels, and a putative correlation between IL-10 and IL-1alpha plasma levels were analysed. Four hundred adult blood samples were obtained from the Finnish Red Cross Blood Transfusion Centre, Tampere. The IL-10, IL-1alpha, IL-1beta and IL-1Ra gene polymorphisms were analysed using PCR. IL-1beta and IL-10 plasma levels were measured using an ELISA method. Our results indicated that increased IL-10 plasma levels were associated with the ATA haplotype (p = 0.03) and, surprisingly, with the IL-1alpha allele 2 carrier status (p = 0.02) in healthy individuals. This IL-1alpha 2+/ATA+ combination was found in 93 subjects out of 400 analysed (23%) and was associated with significantly high IL-10 plasma levels (p = 0.002). When individuals were classified into three groups, with no detectable IL-10 plasma levels (n = 145), with moderate levels (n = 152) and with high levels (n = 100) of IL-10, the IL-1alpha2+/ATA+ combination was more likely present among those with high levels than among those with undetectable levels of IL-10 (OR = 3.3, 95% CI 1.8 - 6.0, p < 0.001) or those with moderate levels of IL-10 (OR = 2.0, 95% CI 1.2 - 3.6, p = 0.012). Besides the observed association between IL-1alpha genotype and IL-10 levels, a moderate correlation was found between IL-10 and IL-1beta levels (r = 0.6, p = 0.01) among IL-10 producers (n = 252). The present findings suggest that the genotype combination of IL-1alpha 2+/ATA+ has a regulatory effect on basal IL-10 levels and that among individuals with measurable IL-10 plasma levels, IL-1beta and IL-10 basal levels correlate. Until now, data on the feedback loop between IL-1 and IL-10 cytokines have been based on studies in vitro, but now our results suggest that this relationship may also exist in vivo.