The aim of this study was to determine whether improved compliance by intensive medication counselling, given by a pharmacist, to elderly patients with chronic stable heart failure can influence both objective and subjective measures of heart failure. Elderly patients were randomly allocated to receive a 3-month counselling programme, or no counselling. Measures recorded at the beginning and end of the study included; submaximal 6-minute exercise tests, visual analogue scores of breathlessness, the Nottingham Health Profile, and clinical signs of heart failure. Compliance was measured by a tablet count and medication knowledge assessed by means of a questionnaire. There was no significant difference between the groups in their initial level of compliance, medication knowledge or other assessments. Compliance improved for the counselled group by 32% (P < 0.001) but remained unchanged for controls. Medication knowledge improved for the counselled group only. The 6-minute exercise test improved by 20 metres from a baseline of 137 m for the counselled group (P < 0.005) but worsened by 22 m for the control group (P < 0.01). Distance to breathlessness improved for the counselled patients and worsened for controls. In contrast bodyweight, jugular venous pressure and Nottingham Health Profile scores did not change significantly for either group. Peripheral and pulmonary oedema scores improved for the counselled group (P < 0.01) but remained unchanged for controls. A small improvement was seen in the visual analogue scores (P < 0.05) for the counselled group only. Improved compliance due to intensive medication counselling had a small but measurable beneficial effect on objective measures of heart failure.(ABSTRACT TRUNCATED AT 250 WORDS)