Compounds used in pharmaceutical, nutritional and cosmetic should be non-toxic. One of the toxicity tests is performed using Artemia spp. as sensitive organisms to various toxins. In these tests, various characteristics of these larvae are investigated. One of the most commonly used factors is evaluation of the survival and death rate of larvae. Detection of live and dead larvae is based on the motility of larvae. However, the rapid movement of larvae makes it difficult to count, and thus the reliability of the test decreases. This study aimed to develop a program for the automatic differentiation of live and dead larvae and their counting in a short time and with high precision. In this regard, a program was designed using image processing rotes to distinguish larvae in MATLAB software. Motility of larvae was detected by image subtraction. The program receives images sent from the camera at two time range (20 seconds) at the end of incubation period and then processes. At each interval, the background is calculated for the same interval. This will cause larvae that have died so far to be added to the background and not counted in the detection of live larvae. Eventually, the program could accurately count live and dead larvae. There is no significant difference between findings from the program and the findings provided by the expert user (P>0.05). Benefits of this improved test with programming are lack of need to aseptic conditions, costly materials and equipment and user presence as well as reduced errors