Using improved microscopy of urine sediment, the consistent finding of bacteria, usually gram-positive cocci at low counts, free or even in casts, in the sediment of carefully collected urines from patients with systemic illnesses has led to the need to reconsider their exclusion by the customary criteria for ‘significance’ of bacteria in urine. Since ‘significance’ is currently based upon mathematical assumptions limited to high counts ( > 105 colony-forming units/ ml) for the prediction of clinical pyelonephritis alone, a digital computer program was created to predict the full spectrum of the expected concentration of bacteria in bladder urine versus time for a very wide range of possible bacterial division times, bladder kinetics and urine flow rates. Curves generated resulted in the discovery of very simple rules based on an easily calculated discriminant, the host’s critical division time (CDT) for any bacterial species in his urine. (1) If the division time in the urine of a species entering the urinary tract is shorter than the CDT, the bacteria will proliferate to > 105 cfu/ml. Published data on growth in human urine show that very few bacterial species can divide so fast in urine, and those are the ones currently considered ‘significant’. Except for some enterococci, streptococci cannot. (2) With a division time only marginally longer than the CDT, any bacterium would wash out unless continuously supplemented via the kidney or from the bladder wall. (3) With a continued supplement and the longer division time, the concentration would fall to a low plateau, and that plateau is diagnostic of a continued supplement. The cocci observed by microscopy are fastidious or dead. They grow poorly if at all in urine, and thus are not likely to ascend the urinary tract. Their appearance corresponds to the earlier studies of bacteriuria and to the known excretion of blood-borne bacteria in natural disease, whether or not there are anatomical changes in the kidney. It is suggested that the low-level coccal bacteriuria found is a marker for scant bacteremia in many systemic diseases for which a bacterial provocation has been sought.