Highly sensitive calorimetry was evaluated for its potential for the monitoring of the bacterial contamination of drinking water. For that purpose, water samples were added to different media and bacterial growth was followed microcalorimetrically. Drinking water samples were either tested untreated or after artificial contamination with selected bacterial strains. Two bacteria, two growth media and two growth conditions (i.e. aerobic and anaerobic) were applied. Even minor initial bacterial contamination (1–100 cells mL⁻¹) gave rise to calorimetric signals after 5–17 h growth. Calorimetry was thus faster than detection of bacterial contamination by plating on agar and inspection of colony formation, which was performed for comparison. It was further demonstrated theoretically that calorimetric detection is superior to colony detection. The heat production curves were characteristic for the strain, the medium and the growth conditions. It is hypothesized that a further refinement of the microcalorimetric method via the application of sets of specific media in combination with selective growth conditions should allow delimiting the identity of the contaminant.