This paper deals with 1) Analysis of the agar droplet technique for viable bacterial counts with respect to a new technical aid (Colworth DROPLETTETM); 2) Comparison of the results obtained from two conventional plate count techniques and the agar-droplet method; 3) Examples for the practical use of this new technique. As a helpful aid for the agar-droplet technique there now exists an apparatus (Colworth DROPLETTETM, Fig.1) which facilitates dilution and dispension of bacteria containing agar, as well as counting of colonies grown in agar droplets. Precision and reproducibility of the diluter/dispenser was found to be within the limits of the usual sampling error (Tab.1, Fig. 2, Fig. 3). The visibility of colonies grown in agar droplets is improved by a ten-fold enlargement on the ground glass screen of the viewer (Fig. 7). In order to obtain droplets of equal size a standardised dropping-technique is required. The fluid agar must not exceed a constant temperature of 50 degrees C as otherwise four of the five species tested (Staph. aureus, E. coli, Kl. aerogenes, Ps. aeruginosa and Enterococci) were significantly reduced in number during the manipulation (Fig. 4, Fig. 5). Agar droplet technique, flooding technique and pour plates gave similar results with gram-positive and gram-negative bacteria from medical routine materials. Comparing the arithmetic means, viable counts of Staph. aureus were slightly higher from pour plates while E. coli and Kl. aerogenes gave the highest counts on flooded plates (Tab. 8-11). All of the three methods produced relatively too low counts when the number of bacteria per sample surmounted 300. Counts below 10 allow only a very poor estimation of the actual number of bacteria (Tab. 12). The variability of droplet counts with bacterial numbers higher than 25 was less then the one of corresponding plate counts (Tab. 14-16). Significant savings in materials, labour and incubator space are made with the agar droplet method. As examples for the practical use of the new method the results of the kinetics of thermally inactivated as well as of growing bacteria are presented.