AbstractLet $H:\text {dom}(H)\subseteq \mathfrak {F}\to \mathfrak {F}$ H : dom ( H ) ⊆ F → F be self-adjoint and let $A:\text {dom}(H)\to \mathfrak {F}$ A : dom ( H ) → F (playing the role of the annihilation operator) be H-bounded. Assuming some additional hypotheses on A (so that the creation operator A∗ is a singular perturbation of H), by a twofold application of a resolvent Kreı̆n-type formula, we build self-adjoint realizations $\widehat H$ H ̂ of the formal Hamiltonian H + A∗ + A with $\text {dom}(H)\cap \text {dom}(\widehat H)=\{0\}$ dom ( H ) ∩ dom ( H ̂ ) = { 0 } . We give an explicit characterization of $\text {dom}(\widehat H)$ dom ( H ̂ ) and provide a formula for the resolvent difference $(-\widehat H+z)^{-1}-(-H+z)^{-1}$ ( − H ̂ + z ) − 1 − ( − H + z ) − 1 . Moreover, we consider the problem of the description of $\widehat H$ H ̂ as a (norm resolvent) limit of sequences of the kind $H+A^{*}_{n}+A_{n}+E_{n}$ H + A n ∗ + A n + E n , where the An’s are regularized operators approximating A and the En’s are suitable renormalizing bounded operators. These results show the connection between the construction of singular perturbations of self-adjoint operators by Kreı̆n’s resolvent formula and nonperturbative theory of renormalizable models in Quantum Field Theory; in particular, as an explicit example, we consider the Nelson model.