BACKGROUND: Increasing evidence shows that Hepatitis B virus infection associates with B-cell but not T-cell malignancies. It remains unclear (a) whether this association is restricted to discrete subtypes of B-cell neoplasms and (b) if occult hepatitis B virus infection can impact on the risk of B-cell malignancy. METHODS: We analysed the prevalence of occult hepatitis B virus infection in three age and sex matched groups: patients with multiple myeloma, chronic lymphocytic leukaemia and healthy volunteers (N=80 each group). Hepatitis B virus sequences were detected by PCR in blood mononuclear cells isolated prior to treatment. RESULTS: Fifteen subjects tested positive for occult hepatitis B virus infection and its distribution significantly favoured chronic lymphocytic leukaemia (p<0.02) over the other groups. No difference in age, gender and proportion of anti-HBc seropositivity was noted according to occult hepatitis B virus infection status. Chronic lymphocytic leukaemia had an odds ratio of 4.6 (95% CI 1.5-13.9) for the presence of occult hepatitis B virus infection in comparison to multiple myeloma and controls. Most occult hepatitis B virus infection cases (10/15, 67%) were detected in completely hepatitis B virus seronegative subjects. CONCLUSIONS: Our data support a potentially causal relationship for hepatitis B virus in chronic lymphocytic leukaemia but not in multiple myeloma. HBsAg seropositivity alone may bias the study of this association, potentially leading to underestimation. Primary occult hepatitis B virus infection appears the most frequent setting in our patients, extending the clinical relevance of hepatitis B virus vaccination to a preventative measure for B-cell neoplasms.