Early screening of Alicyclobacillus spp. in fruit juices is a major applicative goal for the food industry, since juice contamination can lead to considerable loss of quality, and subsequently, to economic damages for juice producers. This paper presents an accurate study to assess and confirm the EOS507 electronic nose's (EN) ability of diagnosing Alicyclobacillus acidoterrestris spoilage in artificially contaminated fruit juices. The authors experimental results have shown that the EOS507 can early identify, just after 24 hours from inoculation, contaminated orange and pear juices with an excellent classification rate close to 90% and with a detection threshold as low as 103 cfu/ml. In apple juice the detection threshold was about 105 cfu/ml, thus requiring longer incubation times (72 hours). PLS regression of EOS507 data can be also used to predict with fair accuracy the colony-forming units concentration of the bacteria. These results were supported by the GC/MS/MS measurements of specific chemical markers, such as guaiacol.