A discussion of nitrosative deamination of cytosine 1 is presented that argues for the formation of 6 by diazotization of 1 to cytosinediazonium ion 2 and its electrostatic complex 3, dediazoniation to 4 ↔ 5, and amide-bond cleavage to 6. The reaction channels available to 6 include hydrolytic deglycation to 3-isocyanatoacrylonitrile 7, water addition to carbamic acid 9 with the possibility for re-closure to uracil 13, water addition to carbamic acid 9, and decarboxylation to 3-aminoacrylonitrile 10. With a view to the instability of the carbamic acid 9, the carbamate models ethyl (Z)-2-cyanovinylcarbamate 14 and (Z)-2-cyano-1-tert-butylvinylcarbamate 20 were studied. Acid-catalyzed hydrolysis of 14 leads to 2-amino- carbonylphenylcarbamate 15, and its cyclization yields the benzo-fused uracil quinazoline-2,4-dione 16. In contrast to the aromatic system 14, acid-catalyzed cyclization cannot compete with oligomerization in the case of 20, and 5-tert-butyluracil 22 is accessible only with base-catalysis. It is shown that 23, the parent of 10, also easily polymerizes. The experimental results provide a rationale as to why 9, 10, and 12 would have escaped detection in in vitro studies: they would have oligomerized. In contrast to the in vitro experiments, the oligomerizations of 9, 10, or 12 clearly are not relevant in vivo because of low monomer concentrations. With the exclusion of recyclization and of oligomerization in vivo, attention thus needs to focus on (Z)-3- aminoacrylonitrile 10 as the most likely deamination product of cytosine aside from uracil.