Understanding yield-limiting traits can help researchers reduce rice yield gap, a key step in increasing yield and sustainability. This research therefore involved two steps: (i) monitoring 100 paddy fields of farmers to identify the most important variables to enter into a CPA model (comparative performance analysis) for the yield gap in the Sari region of northern Iran in 2015 to 2016, and (ii) investigating the effect of seedling age on local and improved cultivars in this region (Mazandaran province) in 2016 and 2017. The field experiment was conducted in a factorial experiment based on a randomized complete block design (RCBD) with three replications. The three seedling ages were 20, 27 and 34 days old for six Iranian rice cultivars (three local cultivars: Tarom Hashemi, Tarom Mahalli and Sang Tarom; and three improved rice cultivars: Fajr, Neda and Shiroodi). Of the 150 variables studied in the first experiment, eight independent variables were chosen for the final model. In the yield model, the average and maximum yields were 4437 and 6690 kg ha-1 , respectively, with an estimated yield gap of 2253 kg ha-1 . This yield gap was related to seedling age, transplanting date and seedling number per hill. Those factors were responsible for 104, 224 and 534 kg ha-1 , i.e. 5, 10 and 24 percent, respectively. The results of the second experiment demonstrated that all investigated traits were significantly affected by seedling age. Twenty-day-old seedlings yielded the greatest panicle length, flag leaf length, number of panicles per m2 , filled spikelet percentage, grain yield, harvest index, protein content, protein yield, nitrogen harvest index. Accordingly, the model’s precision is good and can be applied both to estimate the quantity of yield gap and to determine the portion of each constraint in the yield variables. Importantly, as the calculated yield potential is reached based on actual data in each paddy field, the yield potential is attainable. A seedling age of 20 days created better conditions for growth and yield for both local and improved cultivars.