There is wide consensus that lymphocyte predominance Hodgkin's disease (LPHD) represents a distinct clinicopathological entity of B-cell origin. However, inconsistent results of immunophenotyping studies and low confirmation rates among multi-center trials pose the question of whether LPHD really expresses heterogeneous marker profiles or whether it represents a mixture of morphologically similar entities. Among 2,836 cases reviewed by the German Hodgkin Study Group, immunophenotyping was performed on 1) cases classified or confirmed as LPHD by the reference panel (n = 104) or 2) cases not confirmed as LPHD but classified as classical HD (cHD) within the reference study trial (n = 104). In most cases, immunohistochemistry revealed a phenotype either LPHD-like (CD20+, CD15-, CD30-, CD45+) or cHD-like (CD15+, CD30+, CD20-, CD45-). In 27 cases, the immunophenotype was not fully conclusive. Additional markers for Epstein-Barr virus and CD57 and in situ hybridization for mRNA light chains allowed for a more clear-cut distinction between LPHD and cHD. However, in 25 of 104 cases, immunohistochemistry disproved the morphological diagnosis of LPHD of the panel experts, whereas 13 cases originally not confirmed as LPHD showed a LPHD-like immunopattern. Immunohistochemically confirmed LPHD cases showed a significantly better freedom from treatment failure (P = 0.033) than cHD; this was not observed in the original study classification based only on morphology (P > 0.05). Significantly better survival for LPHD cases improved from P = 0.047 (original study classification) to P = 0.0071 when classified by immunohistochemistry. Our results show that LPHD is a more immunohistochemical rather than a purely morphological diagnosis. Immunophenotyping of HD biopsies suspected of being LPHD is mandatory when a modified therapy protocol, that is, one different from those used in cHD, is discussed.