Three water soluble copolymers based on N-(2-hydroxypropyl)methacrylamide were prepared. Copolymer I contains adriamycin, a chemotherapeutic agent, attached via enzymatically degradable oligopeptide (glycylphenylalanylleucylglycine; G-F-L-G) side chains. The other two copolymers contained the photosensitizer, meso-chlorin e6 monoethylene diamine disodium salt (Mce6). In Copolymer II, the chlorin is attached via the degradable G-F-L-G sequence, and it was bound by the nondegradable glycyl spacer in Copolymer III. Initially, the copolymers were characterized separately in vitro and in vivo. Combinations of the copolymer bound chemotherapeutic agent and each of the copolymer bound photosensitizers were then assessed for antitumor effect in vivo. Localization/retention studies (A/J mice; Neuro 2A neuroblastoma solid tumor) were performed with the two copolymers containing Mce6 as well as the free drug. Results of these experiments demonstrated a very different tumor uptake profile for the two copolymers. While the free drug was rapidly cleared from tumor tissue, the copolymer containing Mce6 attached via the non-degradable bond was retained for an extended period; drug concentrations in the tumor were high even after 5 days. On the other hand, a high concentration of the copolymer containing Mce6 bound via the degradable sequence was taken up by the tumor, yet its concentration in the tumor was substantially diminished at 48 h after administration. This shows indirect evidence of in vivo cleavage of Mce6 from the copolymer in the lysosomal compartment which is supported by direct evidence of cleavage by cathepsin B (a lysosomal enzyme) in vitro. Antitumor effects were assessed on Neuro 2A neuroblastoma induced in A/J mice for all three copolymers. Photodynamic therapy (PDT) proved the copolymer with Mce6 bound via the degradable oligopeptide sequence to be a more effective photosensitizer in vivo than the other chlorin containing copolymer. The difference in activity was consistent with the results obtained by photophysical analyses in which the free drug had a higher quantum yield of singlet oxygen generation than the polymer bound drug in buffer. The quantum yield of singlet oxygen generation increased with the enzymatic cleavage of the chlorin from the copolymer. Conditions were subsequently determined for which chemotherapy or PDT would show some antitumor effect, yet be incapable of curing tumors. Finally, combination therapy experiments were performed in which the copolymer bound adriamycin was mixed with either of the copolymer bound chlorin compounds and injected intravenously (i.v.) into the tail veins of mice.(ABSTRACT TRUNCATED AT 400 WORDS)