Many patients with cancer develop heart failure (HF), and many patients with HF develop cancer. Inter-relationships between their natural histories are rarely reported. Health records were obtained for a Scottish region. People aged >50 years were classified by the presence or absence of HF, loop diuretics (a pharmacological marker of congestion), and cancer. Incident cancer, mortality, and cause of death were recorded. Of 317 178 people aged >50 years, 11 268 (3.6%) had HF of whom 6276 were prescribed loop diuretics; a further 19 044 (6.0%) received loop diuretics. Thus, 30 312 (9.6%) people met an expanded definition of heart failure (HFexp). Annual cancer incidence was slightly higher for those with HFexp (incidence rate ratio 1.10; P < .001), varying amongst cancer types. For people with neither cancer nor HFexp, mortality was <3% annually; most deaths were cardiovascular. For those with cancer but not HFexp, annual mortality was 6.3% for women and 9.0% for men; most died of cancer. For those with HFexp but not cancer, annual mortality ranged from 6.6% to 18.4% depending on sex and HFexp criteria applied; most deaths were cardiovascular. For those with both cancer and HFexp, annual mortality ranged from 14.5% to 28.4%; cardiovascular and cancer mortality rates were similar. Few patients died at home (∼20%). Patients with HFexp have a slightly higher risk of cancer overall, but this may vary according to cancer type. Patients with both cancer and HFexp have a poor prognosis with a similar proportion of deaths attributed to cancer and cardiovascular disease.