Mutagenesis is commonly employed for strain engineering using contrasting site-specific and untargeted approaches. Site-specific approaches require a high level of knowledge at both the DNA and protein levels to implement highly specified genetic changes, while untargeted approaches require more extensive characterization after a desired mutant phenotype is screened or selected from a pool of randomly generated mutants. Localized mutagenesis techniques, such as error-prone PCR, have been developed to confine mutation to a specific region of DNA and offer an attractive compromise between untargeted and site-specific methods. However, inefficiencies in cloning, transformation, and library construction can make such approaches time consuming. We have developed a system for tunable mutagenesis localized to plasmid vectors based on the controlled overexpression of error-prone DNA polymerase IV (Pol IV) in growing Escherichia coli cells. We showed that mutation is elevated and confined to a target vector in our system due to co-expression of a DNA nickase and placement of its recognition site on target vectors. We provide evidence that this system's mechanism involves error-prone replication by Pol IV during the homologous recombination repair process, a mechanism previously proposed for stationary phase Pol IV-dependent mutation. In addition to providing mechanistic understanding, experiments with our system in diverse strain backgrounds helped to identify polB and mutS deletion backgrounds in which mutation was elevated, and confirmed the requirement for a toxin-antitoxin system encoded by the yafNOP genes, located downstream of chromosomal dinB, encoding Pol IV. We also examined the effects of different numbers of nickase recognition sites and the addition of chi sites to our target vectors, leading to increased mutation rates. Finally, we employed another localized mutagenesis technique to optimize our Pol IV and nickase overexpression system, showing that optimal mutation rates are achieved at lower copy numbers when the expression system is plasmid-borne. In the course of designing and improving our system, we developed new techniques to support mutation research. These include a method for placing confidence intervals on ratios of mutation rate estimates using a bootstrapping approach and a high-throughput microplate fluctuation test inspired by fluorescence Miller assays and the Ames II microplate mutagenicity test.