In an initial study of 20 fresh ovarian tumour samples, we compared the immunohistochemical positivity of staining of the p53 protein with the presence of missense mutations of the p53 gene. This revealed a prevalence of 50% with a perfect correlation between mutation and immunohistochemical staining. Detection of the p53 protein by immunohistochemistry was, therefore, used as a reliable indicator for the presence of P53 mutation, and was applied to a study of an archival series of 93 ovarian tumours. Positive immunostaining of the p53 protein was observed in 47% of this series. Cox regression was used to assess whether various clinical variables and P53 mutation were related to survival. As a result, it was found that positive staining of the p53 protein was independent of age, tumour differentiation, tumour type, though possibly not stage. There was some evidence that p53 positivity was associated with reduced survival after adjusting for other variables, but the result was not statistically significant.