Broth dilution minimal inhibitory concentration (MIC) readings were compared after different incubation periods and with different inoculum concentrations. The purpose was to determine the best conditions for obtaining early results as close as possible to overnight readings. Initially, 76 antibiotic-organism combinations were tested using the International Collaborative Study technique and inoculum and were read after 3, 8, and 18 h of incubation. Approximately 28% of tests showed fourfold or greater increases in MICs after 18 h of incubation compared with the 3-h readings. No overnight MICs were lower than early readings. MICs of single antibiotics against seven organisms were also read with an automatic particle counter to confirm the validity of the visual readings. Experiments were made to determine whether inoculum manipulation could reconcile the differences between 3-and 18-h MIC results. One hundred and eight organism-antibiotic combinations were tested comparing 3-h MIC readings using an inoculum of 107 organisms per ml with overnight readings using 105 per ml. In 71 cases, readings with both inocula were within the range tested and 57 (86%) were within i 1 log2 of each other and followed an approximately normal distribution. Improved comparability between early read and overnight MICs thus may be achieved by inoculum manipulation, and this may be a suitable approach in the future development of automated procedures.