In most pregnant women, blood vessels, particularly those supplying the uterus, dilate, contributing to decreased total peripheral resistance and hypotension. However, women with pre-eclampsia are hypertensive. The placenta, and sometimes other maternal organs, show reduced blood flow, which has been linked to constriction of the vasculature. Since nerves control blood vessel diameter, changes in innervation may be critical for normal vascular adaptations to pregnancy. As a first step towards identifying differences in blood vessel innervation in pre-eclamptic pregnancies, we defined how uterine vascular nerves change during normal pregnancy by comparing virgin and 20-day pregnant rats. The rats were anaesthetised with pentobarbitone sodium (60 mg/100 g), perfused with phosphate-buffered saline and euthanized. Their uteri were removed, bubbled with calcium-free Krebs solution for 30-60 min, pinned flat and fixed with phosphate-buffered 4% formaldehyde. We prepared whole mount preparations of intact, full-thickness uterine horns, immunostained them and then embedded them in resin. We detected six neurochemicals that stain specific populations of nerves: tyrosine hydroxylase (TH) and neuropeptide Y (NPY) to identify sympathetic nerves, vesicular acetylcholine transporter (VAChT) and nitric oxide synthase (NOS) to identify parasympathetic nerves, and substance P (SP) and calcitonin gene-related peptide (CGRP) to identify sensory nerves. We also visualized neuronal class III β-tubulin, a neurofilament marker that is presumed to label all axons, using the monoclonal antibody TUJ-1. In non-pregnant rats (n = 6), sympathetic, parasympathetic and sensory axons innervated vascular and non-vascular uterine smooth muscle. TH and NPY-immunoreactive axons formed dense plexuses around arterioles; NOS plexuses were moderate; SP, CGRP, VAChT plexuses were sparse. All six types of axons were common in the longitudinal and circular smooth muscle layers of the myometrium. TH-, NPY- and CGRP-containing axons were denser in the linea uteri than in the rest of the non-vascular smooth muscle. In pregnant uterine horns (n = 4-6), the densities of sympathetic (TH and NPY), parasympathetic (NOS and VAChT) and sensory (SP and CGRP) nerves were very substantially reduced. Occasionally, a few axons occurred near blood vessels; but they were usually entirely free of innervation. Immunoreactive axons were also rare in the longitudinal and circular layers of the myometrium. Immunostaining with TUJ-1 also indicated that there was a very significant denervation of all smooth muscle in the pregnant uterus. These results show that pregnancy causes an almost total of loss of sympathetic, parasympathetic and sensory innervation of uterine vascular and non-vascular smooth muscle. Failure of uterine arterioles to denervate completely may contribute to pre-eclampsia.