The potential for pollen-mediated transgene flow into wild or closely related species has provoked unease in terms of transgenic modification of agricultural plant species. One approach to remedy this situation in species whose seeds and fruits are not of particular value is to engineer male sterility into the transgenic lines. In this study, three meiosis-critical genes, namely AHP2, AtRAD51C and SWITCH1 (SWI), were chosen as silencing targets to test the feasibility of incorporating sterility into plants using an RNAi-based approach. Our results indicated that the silencing of each of these genes via hairpin RNA (termed AHPi, RAD51Ci and SWIi lines) in Arabidopsis thaliana yielded a proportion of transgenic plants exhibiting a similar 'partially sterile' phenotype in which less than 50% of pollen was viable. In addition, a 'sterile' phenotype was also evident in a minority of RAD51Ci and SWIi, but not AHPi, lines in which plants yielded no seeds and either produced inviable pollen (RAD51Ci lines) or displayed a complete absence of pollen (SWIi lines). This suggests that AtRAD51C and SWI may function at distinct stages of meiosis. Further analyses of SWIi lines demonstrated that the 'sterile' phenotype was associated with a substantial reduction in the level of targeted gene transcript in floral tissues and resulted from sterility of the male, but not female gametes. This work demonstrates that generating male sterility through the silencing of key genes involved in the regulation of meiosis is feasible, and its advantages and potential applications for transgene containment are discussed.