Rapid and accurate detection and identification of microbial pathogens causing urinary tract infections allow prompt and specific treatment. We optimized specimen processing to maximize the limit of detection (LOD) by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) and evaluated the capacity of combination of MALDI-TOF MS and urine analysis (UA) for direct detection and identification of bacterial pathogens from urine samples. The optimal volumes of processed urine, formic acid/acetonitrile, and supernatant spotted onto the target plate were 15 ml, 3 mu l, and 3 mu l, respectively, yielding a LOD of 1.0 X 10(5) CFU/ml. Among a total of 1,167 urine specimens collected from three hospital centers, 612 (52.4%) and 351 (30.1%) were, respectively, positive by UA and urine culture. Compared with a reference method comprised of urine culture and 16S rRNA gene sequencing, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of MALDI-TOF MS alone and MALDI-TOF MS coupled with UA were 86.6% versus 93.4% (chi(2) = 8.93; P < 0.01), 91.5% versus 96.3% (chi(2) = 7.06; P < 0.01), 81.5% versus 96.4% (chi(2) = 37.32; P < 0.01), and 94.1% versus 93.1% (chi(2) = 0.40; P > 0.05), respectively. No significant performance differences were revealed among the three sites, while specificity and NPV of MALDI-TOF MS for males were significantly higher than those for females (specificity, 94.3% versus 77.3%, chi(2) = 44.90, P < 0.01; NPV, 95.5% versus 86.1%, chi(2) = 18.85, P < 0.01). Our results indicated that the optimization of specimen processing significantly enhanced analytical sensitivity and that the combination of UA and MALDI-TOF MS provided an accurate and rapid detection and identification of bacterial pathogens directly from urine.