Objectives: A handheld biosensor for measuring salivary α-amylase (sAA) was developed for convenient on-site measurement. Previous studies reported some discrepancies in sAA levels measured with a biosensor and a standard assay. This study aimed to compare sAA levels measured with three different methods and the factors affecting its levels. Methods: Thirty-eight participants collected saliva two times for three measurements. First, the collector strip was placed under the tongue for 2 minutes, then the strip was used to measure sAA level on-site immediately (intraoral biosensor; method 1). Then, a participant pooled the saliva for 4 minutes and collected the saliva into the tube which was aliquoted to measure in a laboratory with a handheld biosensor (extraoral biosensor; method 2) and with a standard enzyme kinetic assay (EKA; method 3). Additional experiments were carried out to compare the levels of sAA measured with differences in pooling time and positioning of the collector strip. Results: A high correlation of sAA levels between an extraoral and an EKA measurement (r = 0.989) was observed, while sAA levels measured with an intraoral method showed a significant but weaker correlation with either an EKA (r = 0.475) or an extraoral method (r = 0.436). Saliva pooling time and positioning of the collector strip significantly affected sAA levels. Conclusions: A handheld biosensor is valid to measure sAA levels extraorally. For an intraoral measurement, pooling time and positioning of the collector strip need to be taken into account.