Risks in agriculture and the risk of climate change are increasingly impacting agriculture, and farmers are aware of climate change. Besides production risk, farmers note income variability on their farms. Previous research stated that income variability is seen through the years among EU farmers. In Croatia, average income increased from 2014 to 2017, and higher-income variability was recorded for all farms and horticultural farms. The highest income variability was recorded among wine farms in Croatia (Čop and Njavro, 2022a). To cope with risks, farmers can use on-farm strategies (such as irrigation, changing the harvest date, organic production, etc.) and risk transfer strategies (agricultural insurance, hedging, production contracts, etc.). However, agricultural insurance in Croatia is applied only to 4% of farmers. The main reasons for the low rate of interest for agricultural insurance are high premiums, insurance coverage, and many administration (Smrkulj and Njavro, 2016). Within the Common Agricultural Policy framework, the importance of risk management in agriculture has been recognized, and three risk management measures have been developed for EU farmers. The three measures are subsidized agricultural insurance, an income stabilization tool (IST), and a mutual fund (EU Regulation 1305/2013 and EU Regulation 2393/2017). The income stabilization tool compensates farmers for a severe drop in their income to 70% loss. The IST will be activated when the drop in income exceeds 30 % or (20%) of the average annual income of the individual farmer in the preceding three-year period or a three-year average based on the preceding five-year period excluding the highest and lowest entry (EU Regulation 1305/2013 and EU Regulation 2393/2017). The main advantage of the IST is that IST represents whole farm insurance that covers all risks at the farm, from production and price to income risk. The paper's objectives are to examine the readiness of vine growers and wine producers (VG-WP) to apply an innovative risk management strategy and whether respondents maximize utility by applying an IST. Additionally, in the research, we assess how the behavioral factors (variables of risk preference and subjective probability of income loss), socio-demographic variables, and farm characteristics affect the choice of income stabilization tool. We used a questionnaire and conducted a survey among VG-WP from Adriatic Croatia to collect data. The reason for choosing Adriatic Croatia is a higher exposure of Mediterrane to climate change (Jacobs et al., 2019; Carraro and Sgobbi, 2008; Giorgi and Lionello, 2008). Methods used in the paper are the discrete choice experiment (DCE), the multiple price list (MPL), and the scoring method. Monte Carlo simulation was used to evaluate the impact of IST on the vine business. The DCE examines respondents' stated preferences for innovative risk management strategy through hypothetical choices using random utility theory (McFadden, 1974). VG-WP could choose among alternatives, income stabilization tool (IST) or optout alternative (meaning not choosing IST). IST was defined by four attributes and associated levels: income threshold (10%, 20%, and 30%), compensation (60%, 70%, and 80%), premium (2.000 HRK/HA, 2.500 HRK/HA, 3.200 HRK/HA, 4.200 HRK/HA and 5.400 HRK/HA) and time (3 or 5 years – the time frame used to calculate whether the income drop has reached the threshold). VG-WP received 12 choice cards, and in every one of them, they needed to choose only one alternative with whom they maximize their utility. For choice card design, we used Ngene software, and for model estimation, STATA v. 16. In our modelling approach, we assume that non-price attributes are distributed according to a normal distribution with mean and standard deviation to be estimated and that the error term is distributed according to an iid extreme value (Train, 2009). We estimate the multinomial logit model (MNL) and mixed logit model (RP-MNL) that capture unobserved heterogeneity for the non-price attributes. With MPL, we research farmers' risk preferences and, using a scoring method, examine subjective probabilities of income loss in 2020 compared to 2019. The MNL and RP-MNL models show that the VG-WP of the Adriatic Croatia prefers IST, and by choosing IST, they maximize their utility. Increment in compensation level and the calculation of the average income loss based on three years impact VG-WP's maximizing utility. Premium as an attribute of IST has a negative coefficient that shows that respondents prefer a lower premium. The income threshold is not statistically significant and does not impact VG-WP's IST choice. The MPL and scoring method show that VG-WPs are riskaverse and do not expect high-income losses. Around 85% of respondents expect income loss from 0-50%, while only 15% of VG-WP expect losses higher and equal to 51%. In addition, young farmers and those who deal with viticulture production only are ready to choose IST. Furthermore, additional analyzes show that VG-WP who are pessimistic about the loss of income choose IST. The Monte Carlo simulations show a positive impact of IST on the vine business. The research has confirmed that VG-WPs are ready to choose an innovative risk management strategy (IST). The conducted research is the first research that evaluates the farmers' acceptability of IST using the discrete choice experiment at the EU level and can serve as a basis for the design of IST in Croatia and broader. The scientific contribution of the research is reflected in the development of the application of discrete choice experiment in decision making. It helps to identify the limitations of this method in assessing the introduction of IST as an innovation for risk management. The research limitations are reflected in the application of IST among VG-WP as a hypothetical case. VG-WP does not have to honestly admit whether or not they understood IST and the method of calculating the compensation concerning different cases of income decline. This potential problem has been reduced by a detailed description of the functioning of the IST during interviewing. An additional limitation of the research may be the geographical distribution of the respondents. Apart from Adriatic Croatia, the research could be extended to the whole of Croatia (and the EU) to take into account the opinions and IST preferences of farmers. The research results can assist policy makers in deciding on the introduction of IST in Croatian agricultural policy and help design IST. Previous research stated that only three countries have planned to introduce IST, and only Italy has included IST in existing risk management instruments in agriculture. Limitations of the application of IST are the lack of legislation and the establishment of IST, the lack of practice examples, and the lack of understanding of the advantages and disadvantages of IST. From our viewpoint, already introduced subsidized insurance can lead to an increment of contracted insurance policies and can reduce farmers' aversion to the introduction, selection, and application of innovative subsidized CAP measures in the future. This doctoral research can encourage further research on experiments and experimental methods, emphasizing DCE in risk management in agriculture and decision making. The research design can serve as a basis for similar research on assessing farmers' preferences according to a mutual fund or new types of insurance (e.g., revenue insurance, index insurance). DCE research can also be applied to the design of the Common Agricultural Policy, i.e., agricultural measures. In addition, an experiment to assess the choice of IST could be conducted in the other EU Member States. In addition to the applied behavioral variables (subjective probabilities and risk preferences), future research may include variables of trust, the moral hazard of respondents, and their willingness to choose IST. The conducted research is the first such research based on IST selection (new Rural Development Programme measures) and the impact of behavioral variables on IST selection.