Commercial carbapenem antibiotics are being used to treat multidrug resistant (MDR) and extensively drug resistant (XDR) tuberculosis. Like other β-lactams, carbapenems are irreversible inhibitors of serine d,d-transpeptidases involved in peptidoglycan biosynthesis. In addition to d,d-transpeptidases, mycobacteria also utilize nonhomologous cysteine l,d-transpeptidases (Ldts) to cross-link the stem peptides of peptidoglycan, and carbapenems form long-lived acyl-enzymes with Ldts. Commercial carbapenems are C2 modifications of a common scaffold. This study describes the synthesis of a series of atypical, C5α modifications of the carbapenem scaffold, microbiological evaluation against Mycobacterium tuberculosis (Mtb) and the nontuberculous mycobacterial species, Mycobacterium abscessus (Mab), as well as acylation of an important mycobacterial target Ldt, LdtMt2. In vitro evaluation of these C5α-modified carbapenems revealed compounds with standalone (i.e., in the absence of a β-lactamase inhibitor) minimum inhibitory concentrations (MICs) superior to meropenem-clavulanate for Mtb, and meropenem-avibactam for Mab. Time-kill kinetics assays showed better killing (2–4 log decrease) of Mtb and Mab with lower concentrations of compound 10a as compared to meropenem. Although susceptibility of clinical isolates to meropenem varied by nearly 100-fold, 10a maintained excellent activity against all Mtb and Mab strains. High resolution mass spectrometry revealed that 10a acylates LdtMt2 at a rate comparable to meropenem, but subsequently undergoes an unprecedented carbapenem fragmentation, leading to an acyl-enzyme with mass of Δm = +86 Da. Rationale for the divergence of the nonhydrolytic fragmentation of the LdtMt2 acyl-enzymes is proposed. The observed activity illustrates the potential of novel atypical carbapenems as prospective candidates for treatment of Mtb and Mab infections.