In order to improve saffron yield and modify crop pattern in arid areas, a split plot experiment was conducted based on randomized complete block design with three replications in a research field in Charkhab village of Yazd province during 2013-2016. The main plots included; irrigation in three levels (100%, 75% and 50% of ETc), and sub plots were seven fertilizer management including; control (zero fertilization), applying 10 t. ha-1 cow manure(10M), 10 t. ha-1 vermicompost by spreading(10V), 10 t. ha-1 vermicompost buried under planting corm rows (10RV), 5 t. ha-1 vermicompost buried under planting corm rows (5RV), and pouring 10 L. ha-1 Humaster Saffron fertilizer for 1 (SH) or 2 (CH) times after flowering. The results showed that as irrigation volume decreased from 75% to 50% ETc, leaf dry weight, highest corm weight, total corm weight and effective corm number decreased significantly. The differences among irrigation treatments increased year by year, So that, there were no significant difference among treatments at the first year while the flower number (100. 76 flower per plot in average) at 100% irrigation treatment was significantly more than the other two treatments. Also, as irrigation volume declined from 100% to 75% of ETc, total corm number decreased more than 40% while total corm weight diminished less than 30%. Among fertilizer managements, 10RV and CH treatments showed better performance in the second and third years. So that, maximum flower number was observed in 10RV treatment by 71. 78 and 98. 22 flower per plot at the second and third experimental year, respectively, although, the pure gain of CH treatment was about 13. 5 times more than the best treatment (10RV). In general, the results showed that using Humaster-Saffron fertilizer had more relative advantages in comparison with other treatments. However, it is recommended to test more amounts of this fertilizer in future research studies.