Understanding the timing of crop developmental stages is important for farm operational decision-making and also for development of crop simulation models. Temperature and photoperiod can be used to simulate phenology and are surrogates for complex biochemical pathways leading to the appearance of certain developmental stages at the whole plant level. Commonly, in the multiplicative approach, it is assumed that the developmental rates respond to temperature and photoperiod in a linear fashion. Literature, however, suggests that soybean developmental rates respond non-linearly to these abiotic factors. The use of non-linear temperature and photoperiod functions and separation of floral induction and evocation for simulating flowering (R1) were part of the approach used in the development of a new soybean phenology model (SOYDEV). Results based on five years irrigated experiment in Lincoln, NE (1999, 2000, 2001, 2002, and 2004) with a semi-determinate cultivar, NE3001, showed that SOYDEV offers some improvement in prediction of major phenological stages such as unifoliate (V1), flowering (R1), beginning seed (R5), and physiological maturity (R7) as compared to CROPGRO model. We also introduced parameters generalization for indeterminate cultivars with maturity group 3.0 to 3.9 to provide a more “user-friendly” option of the model.