Factors affecting viable cell counts in groundwater or sediments were studied with samples from the Segeberg Forest test area in northern Germany. There was very little variation in results with the season (April, August, November) or depth of sampling; generally there were 10(3)-10(4) aerobic cells per ml or g sediment. Long incubation times resulted in higher cell counts; groundwater samples required 4-5 weeks, and sediment extracts had to be cultured for 7 weeks. Total cell counts in sediment were 10(2)-10(4) cell/g higher than viable cell counts of aerobes. This was explained partly by the additional presence of anaerobes and partly by the observation that some morphotypes may not have grown under our conditions. Viable cell counts were not influenced by cell extraction from the sediment with either Na-pyrophosphate or groundwater extracts. However, iron-precipitating or manganese-oxidizing bacteria were better extracted with sterile groundwater. The microflora of wells was more numerous than that of the free aquifer; consequently it was better to pump off all well water before aquifer water was sampled. The diameter of the well was also important; thinner tubes had higher cell counts than those with wider diameter. For sampling, wells should be at least 1 year old, since young wells contain higher numbers of microorganisms due to underground disturbances from the drilling. Turbid water samples could be clarified by filtration, but this reduced the viable counts by 1-2 orders of magnitude. Two different media inoculated with a sample dilution resulted in the same cell counts, but their microbial diversity was different. Storage of groundwater samples before processing resulted in up to 17-fold increases in cell counts and loss of diversity in the first 24 hours. Cell numbers decreased slowly during longer storage.