Abstract Imaging mass spectrometry (IMS) is a molecular imaging technique for various molecules such as compounds, metabolites and proteins to visualize spatial distribution of these components in a tissue sample. In recent years, the measurement of the spatial distribution of components using IMS technique is being carried out intensively in many fields including pathology and drug discovery. Although the importance of IMS has been recognized widely, there are not so many efficient software programs to handle these kinds of data because of the difficulty in treating huge amount of spectra data from many spots, which sometimes amounts to tens of thousands. In this presentation we introduce new software we developed and report its performance. In order to correspond to various purposes, the software was made to consist of four programs as shown below. They achieve quick, automatic, and comprehensive analyses of the important peaks for the first time in the world. To detect important peaks of bio-molecules in each spectrum from mass spectrometry, we first developed a high-speed program for peak picking using a common peak method for IMS data, named IMS Convolution (IMSC). Once some regions of interests (ROI) such as cancer, interstitial and normal regions in a tissue sample are specified through graphical interface program, common peaks in each ROI are automatically detected by IMS convolution. Users can modify a definition of a threshold of common peaks easily, and the IMSC software picks up revised common peaks very quickly. Furthermore, we developed three more programs, named Spatial Peak Detectors (SPeaD) -1, 2 and 3, which use peak picking results by the IMSC. SPeaD-1 is for detecting a small set of peaks and it can discriminate two or more ROIs based on a method from machine learning theory. SPeaD-2 is for detecting cell specific peaks without ROI information. Cancer specific small molecules can be detected when cancer cells are distributed individually in normal tissue. SPeaD-3 is for detecting region specific peaks without ROI information. SPeaD-3 picks up all peaks of the same mass expressed in a clustered region defined by sets of adjacent spots. In evaluating the performance of the software, we used quasi-samples for IMSC and SPeaD-1 and cancer tissue samples for SPeaD-3. IMSC detects all the peaks beyond the noise level. Computation time is within 10 minutes for the data of ANALYZE format file size about 2.5GB with 62,500 (= 250 x 250) spots and with the range of m/z 650-1500. A workstation used for the analysis is equipped with Xeon E5504 CPU (2.0GHz) and 3GB memory. For SPeaD programs, computation time depends on the number of spots and m/z range but falls within about 30 minutes. Furthermore, we analyzed IMS data for hepatic micrometastasis of human colon cancer xenografts in superimmunodeficient NOG mice. Some specific bio-molecules of hepatic micrometastasis were detected quickly and automatically. Further details are given in the presentation. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3963. doi:1538-7445.AM2012-3963