Diagnostic medical microbiology laboratories detect and identify pathogens in submitted specimens. The techniques used should maximise the detection of pathogens (sensitivity) while minimising the number of tests for their detection (efficiency). To achieve the best compromise between sensitivity and efficiency, it is necessary to have information on both the relative prevalence and clinical importance of various pathogens within the relevant community, and the relative efficiency of various detection techniques. This investigation had three primary objectives: to establish what pathogens were associated with community-acquired gastrointestinal symptoms in the Eastern Bay of Plenty, and the incidence and relative importance of each; to compare the merits of various methods for detecting these pathogens (in those cases where more than one method was available); and to collect data from patients so as to identify potential sources and/or risk factors for infection. 997 faecal specimens from 716 episodes of illness were tested over a one year period. Patients completed a questionnaire on symptoms, and food and environmental exposures. Using one or more standard techniques, the specimens were tested for bacteria and parasites which may cause gastroenteritis. Specimens from young children were also tested for the presence of rotavirus. The incidence rates of the various pathogens, expressed as a rate per 100 000 persons per year, were as follows: Blastocystis hominis, 358; Campylobacter species, 208; Giardia lamblia, 158; Yersinia species, 87; Cryptosporidium parvum, 67; Salmonella species, 62; Aeromonas species, 62; Dientamoeba fragilis, 29; Plesiomonas shigelloides, 21; Escherichia coli (E coli) O157, 4; Vibrio cholerae non-O1, non-O139, 4; and Shigella species < 4. Faecal specimen macroscopic form, microscopic findings, season, and patient age showed little correlation with the presence of specific pathogens. Consequently the tests selected for the detection of pathogens in faeces should not be based on any of the above parameters. Furthermore, the symptoms associated with parasitic and bacterial infections were similar, so it is not possible to select the appropriate tests on this basis. The presence of rotavirus in patients older than five years was not investigated so incidence in the general population can not be calculated. A study of all age groups for the presence of this organism would be appropriate. From the above findings, and an evaluation of the literature, it is recommended that all specimens should be examined for the following organisms and, on the basis of our observations, the most cost-effective method is shown in brackets: Salmonella (selenite enrichment subcultured to xylose lysine desoxycholate agar); Shigella (none were detected, so a cost-effective medium could not be determined), Campylobacter (5% sheep blood agar supplemented with 32 mg/1 cefoperazone); Yersinia (Yersinia selective agar (YSA), plus selenite enrichment subcultured to YSA); Giardia lamblia (detection of antigen); Cryptosporidium parvum (detection of antigen). While routine testing for E coli O157 is not recommended, laboratories should have the capability to test for this pathogen if a patient presents with haemolytic uraemic syndrome, thrombotic thrombocytopenic purpura or unexplained bloody diarrhoea. Likewise, routine culture for Vibrio species is not recommended; however, laboratories should test specimens using thiosulphate citrate bilesalt sucrose agar if the requesting clinician suspects cholera, or the patient has a recent history of shellfish consumption. A trichome stain for Dientamoeba fragilis is recommended for patients with chronic gastrointestinal symptoms who are to be investigated for neoplastic and other non-infectious conditions. Pathogenic parasites other than those noted above were not detected. However, since such organisms are isolated in New Zealand, usually in association with overseas travel or institutionalisation, it is recommended that a trichrome stain and a faecal concentration technique should be performed on specimens from all cases of gastroenteritis who have recently travelled overseas or who are institutionalised. Close liaison between the laboratory and the clinician is essential to ensure appropriate selective testing for these less common pathogens. The presence of Blastocystis hominis and Aeromonads should be reported, but the report should note that their pathogenicity is uncertain. Dientamoeba fragilis and Plesiomonas shigelloides are probably pathogenic, but further work is needed to clarify this point. Correlation of data from the questionnaires and the laboratory findings identified the following risk factors: (the relative risk, 95% confidence interval and p-value are shown in the brackets). Campylobacter species: consumption of unpasteurised milk (4.67,2.39 - 9.11, p = <0.001); Salmonella species: overseas travel (7.20, 1.67 - 20.9, p = 0.040), eating a barbecued meal (4.55, 1.37 - 15.12, p = 0.026), eating shellfish (3.80, 1.18 - 12.21, p = 0.032); Yersinia species: consumption of water from a home supply (3.46, 1.32 - 9.10, p = 0.016), handling cattle (4.88, 1.73 - 13.76, p = 0.008), handling sheep (14.80. 4.93 - 44.46, p = 0.001); Giardia lamblia: consumption of unpasteurised milk (3.93, 1.63 - 9.46, p = 0.011), attendance at a day care centre (2.70, 1.17 - 6.27, p = 0.033), handling cattle (3.39, 1.59 - 7.22, p = 0.005), handling horses (5.27, 1.85 - 14.97, p = 0.002); Cryptosporidium parvum: consumption of water from a home supply (5.08, 1.88 - 13.71, p = 0.002), consumption of unboiled water from a natural waterway (3.97, 1.29 - 12.24, p = 0.031), attendance at a day care centre (3.30, 1.06 - 10.22, p = 0.054), handling cattle (5.41, 1.88 - 15.58, p = 0.006), owning a cat (4.50,1.02 - 19.91, p = 0.029); Plesiomonas shigelloides: eating shellfish (13.67, 1.44 - 130.13, p = 0.020); and Dientamoeba fragilis: consumption of unboiled water from a natural waterway (7.46, 1.71 - 32.48, p = 0.019). The risk factors suggest the value of the following precautions to prevent gastrointestinal infection: maintaining a high standard of both personal hygiene (particularly in the rural environment) and environmental hygiene in areas that food is prepared; avoiding consumption of untreated water or unpasteurised milk; cooking animal-derived food thoroughly - especially barbecued food and shellfish; and washing hands thoroughly after animal contact. Persons with diarrhoeal symptoms should take particular care with personal hygiene. Those travelling overseas should be conscious of the risk associated with the consumption of food and water which is not properly cooked or treated. These findings should assist New Zealand laboratories to optimise their approach to the detection of faecal pathogens and should also assist in formulating policy for prevention of infection by enteric pathogens.