Antibiotic resistance is a growing public health concern. One recent antibiotic resistant determinant is New Dehli Metallo-beta-lactamase (NDM-1). This metallo-beta-lactamse confers high resistance to the carbapenem class of antibiotics, considered one of the last lines of treatment against bacterial infections. It has been reported that NDM-1 was created as a fusion between an ancient metallo-beta-lactamase gene and a sequence encoding the first six amino acids of cytoplasmic aminoglycosidase (aphA6). As a result, it contains an unusual signal peptide, with a negative charge at position P2. There is also some inconsistency in the NDM-1 literature, with several proposed signal peptidase I cleavage sites, and one report stating that NDM-1 is a lipoprotein, with the signal peptide cleaved after glycine at position 25. Two signal peptide variants have been reported from clinical isolates; a proline to arginine substitution at position 9 (P9R) and a proline to alanine substitution at position 28 (P28A). These NDM variants and NDM-1 were cloned and purified with an C-terminal his-tag. We found that both P9R and P28A improved secretion compared to NDM-1, as evidenced by less precursor material present in whole-cell lysates, and grew faster in rich media in E. coli compared to NDM-1. Mass spectrometry analysis of the purified material indicated that NDM-1 and P9R result in signal peptide cleavage at position 21 and 22. Interestingly, the P28A variant creates a different signal peptide cleavage site, at position 27. In none of these samples did we observe lipidation of the cysteine at position 26, indicating NDM-1 is most likely not a lipoprotein. Hence, the signal peptide variants studied here show improved secretion and fitness compared to the ancestral NDM-1. These and other signal peptide mutations may occur in the future that improve the ability of the NDM-1 signal peptide to facilitate secretion to the periplasm and increase antibiotic resistance.