Contamination of foods, especially produce, with Salmonella spp. is a major concern for public health. Several methods are available for the detection of Salmonella in produce, but their relative efficiency for detecting Salmonella in commonly consumed vegetables, often associated with outbreaks of food poisoning, needs to be confirmed. In this study, the effectiveness of three molecular methods for detection of Salmonella in six produce matrices was evaluated and compared to the FDA microbiological detection method. Samples of cilantro (coriander leaves), lettuce, parsley, spinach, tomato, and jalapeno pepper were inoculated with Salmonella serovars at two different levels (10(5) and <10(1) CFU/25 g of produce). The inoculated produce was assayed by the FDA Salmonella culture method (Bacteriological Analytical Manual) and by three molecular methods: quantitative real-time PCR (qPCR), quantitative reverse transcriptase real-time PCR (RT-qPCR), and loop-mediated isothermal amplification (LAMP). Comparable results were obtained by these four methods, which all detected as little as 2 CFU of Salmonella cells/25 g of produce. All control samples (not inoculated) were negative by the four methods. RT-qPCR detects only live Salmonella cells, obviating the danger of false-positive results from nonviable cells. False negatives (inhibition of either qPCR or RT-qPCR) were avoided by the use of either a DNA or an RNA amplification internal control (IAC). Compared to the conventional culture method, the qPCR, RT-qPCR, and LAMP assays allowed faster and equally accurate detection of Salmonella spp. in six high-risk produce commodities.