AIMS--To evaluate the ability of five experienced perinatal pathologists to assess placental maturity reliably by histology. METHODS--Twenty four haematoxylin and eosin slides, six each from placentas of 27, 31, 35, and 39 weeks' gestation, were circulated to five pathologists on three separate occasions. The slides were labelled with the correct or incorrect gestational ages. RESULTS--The mean absolute error over all 360 readings was 2.72 weeks. Only 54% of the slides were assessed within two weeks of the correct gestation. Pathologist tended to overestimate younger gestations and underestimate older gestations. Two, and possibly three, pathologist were influenced by the gestational age state on the label. One pathologist, who did not appear to be influenced by the label, was more accurate in diagnosing gestation of the placentas than other colleagues. CONCLUSIONS--Experienced pathologists can have difficulty in assessing the villous maturity of placentas by histology. They can also be influenced by clinical information provided, such as gestational age. Other observer reliability studies must address the issue of the influence of labelled information on observer variation. A difference in maturation would have to be of a six week magnitude to have a chance of being detected by current methods. This may limit the value of the histological diagnosis of placental dysmaturity as a surrogate marker for uteroplacental ischaemia.