One-third of the recent trials and studies on photodynamic therapy (PDT) used the “first-generation photosensitizers”, hematorphyrin (Hp) and its derivatives HpD and Photofrin (PF), which consist of porphyrin dimers and trimers joined by ether linkages. Like its precursors, aggregation is negligible for PF in methanol, but in water at pH 7.4, a decrease in the quantum yield Φ_Δ of singlet oxygen formation is observed with increasing PF concentration and the variations can be quantitatively described in terms of a monomer-dimer equilibrium with quantum yields of Φ_Δ,M = 0.49 and Φ_Δ,D = 0.076 for monomers and dimers respectively. By comparison, the corresponding values ​​for HpD are 0.64 and 0.11 and for Hp 0.74 and 0.12, respectively, and the yields of unassociated species are always identical in methanol and water. The regular decrease in Φ_Δ,M observed by going from Hp (0.74) to HpD (0.64) then PF (0.49) could be attributed to additional (e-v) deactivation of singlet oxygen by the substituents attached to the macrocycle in HpD and PF. Surprisingly, the ratio Φ_Δ,M/Φ_Δ,D  remains constant and reflects the similarity in aggregate interactions in the hematoporphyrin family, and the relatively low values of Φ_Δ,D  for HpD and PF can be essentially attributed to both a substantial decrease in S_Δ, the fraction of triplet quenching by oxygen which produces O₂ (¹Δ_g), and a competition between triplet energy transfer and photoinduced electron (or partial charge) transfer. The observed evolutions and their origin could be of general relevance and concern many families of photosensitizers. Criticisms on variability and inconsistencies in the production of HpD and PF must be reconsidered.