The oestrogen-inducible pS2 protein has previously been associated with good prognosis for breast cancer patients. In 1987-1988 a series of 145 primary breast cancers were examined for pS2 mRNA using northern blots. On recent examination of mortality data, we were unable to find any association between tumour pS2 positivity and patient survival. One patient in 6 died within 5 years of surgery, regardless of pS2 status. In the oestrogen receptor positive/progesterone receptor positive tumour subgroup of patients, we found no evidence of increased survival for pS2-positive tumours. These results do not support use of pS2 as an indicator of increased survival in an average breast cancer patient population.