Background As a model for both multistep and multipathway carcinogenesis, colorectal neoplastic progression provides paradigms for researching both oncogenes and tumor suppressor genes (TSGs). However, the mechanism of colorectal cancer (CRC) is not completely understood, and many genes may be involved in the colorectal carcinogenesis. The purpose of this study was to screen for the potential TSGs on chromosome lq31.1-32.1 in Chinese patients with sporadic colorectal cancer, to explore whether colorectal cancer in the Chinese population has unique genetic alterations and determine whether other putative TSGs exist and contribute to colon carcinogenesis. Methods Six polymorphic microsatellite markers, at a density of approximately one marker in every 1.6 cM, were chosen for refined loss of heterozygosity (LOH) mapping of lq31.1-32.1. Eighty-three colorectal cancer patients' tumor and normal DNA were analyzed via polymerase chain reaction (PCR) for these microsatellite markers. PCR products were eletrophoresed on an ABI 377 DNA sequencer. Genescan 3.1 and Genotype 2.1 software were used for LOH scanning and analysis. On the basis of refined LOH mapping results, we undertook a microarray-based expression screening to identify tumor association genes in 19 of the CRC cases. Results The average LOH frequency of lq31.1-32.1 was 24.41%, with the highest frequency of 36.73% (18/49) at D1S2622, and the lowest of 16.42% (11/67) at D1S412. A minimal region of frequent deletion was located within a 2 cM genomic segment at D1S413-D1S2622. There was no significant association between LOH of any marker in the studied regions and the clinicopathological data (patient sex, age, tumor size, growth pattern, or Dukes stage). On the basis of refined mapping results, we chose 25 genes located in the D1S413-D1S2622 (lq31.3-32.1) region and presented a microarray-based high throughput screening approach in 19 sporadic CRC cases to identify candidate CRC related tumor suppressor genes. This study found 4 significantly down-expressed genes, including CSRP1, LMOD1, PPP1R12B and CFHL3. There was no significant association between expression levels of CFHL3, CSRP1, LMOD1, PPPIR12B and the clinicopathological data. By database searching, CSRP1 was hypothesized to be a colorectal cancer related tumor suppressor gene. Conclusions Through detailed deletion mapping, we found that the lq31.3-32.1 region might harbor one or more colorectal cancer related tumor suppressor gene(s). And by microarray-based high-throughput screening of candidate genes located in this region and by subsequent database searching, we present the first evidence that CSRP1 might be involved in the progression of CRC.