Three vegetable production districts, namely Mymensingh, Rajshahi and Comilla were selected as case study areas. These areas are the well-known vegetable production area in Bangladesh. In the year 2010-2011, total production area of summer vegetables were 7,467.6 and 6355.0 hectares in Mymensingh and Comilla, 12,131.9 hectares were cultivated in Rajshahi respectively. Total production area of winter vegetables in Mymensingh and Comilla were 5,546.1 and 10,119.0 hectares and 10,220.2 hectares were cultivated in Rajshahi (Yearbook of Agricultural Statistics of Bangladesh, 2011). The difference in production area conditions makes the research more suitable. A total number of 87 farmers taking 27 from Mymensingh, 30 from Comilla, 30 from Rajshahi were randomly selected for interview. A semi-structured pre-tested interview schedule was used for collecting data and information from the bottle gourd farmers during May to July 2014. The revenues and costs of bottle gourd cultivation were calculated for four months. We used total variable cost, total fixed cost, total cost, total revenues, gross farm income, net farm income, BCR over variable cost and total cost to calculate profitability. Total variable cost is the sum of seedling, power tiller, hired labor, fertilizer, pesticides, irrigation and bamboo cost. Fixed cost constituted by family labor cost, interest on operating capital and land rent. Total cost is the sum of total variable and fixed cost. Gross farm is the outcome from total variable cost deducted from total revenue. Net farm income is the result from total revenue minus total cost. BCR over variable cost is the ratio of total revenue and total variable cost and BCR over total cost is the ratio of total revenue and total cost. The functional form of the Cobb-Douglas production function model is given below: Y = AX1b1X2b2 ..................Xnbneui The production function was converted to logarithmic form so that it could be solved by least square method i. e. ln Y= a + b1 ln X1 +b2lnX2 … …………… + bn ln Xn + Ui. The empirical production function model is as follows: lnY = a + b1 ln X1 + b2 lnX2 + b3 lnX3 + b4 lnX4 + b5 lnX5 + b6 lnX6 + b7 lnX7 +b8 lnX8 + b9 lnX9 + b10 lnX10 + b11 lnX11 + b12 lnX12+ Ui. Where, Y = Yield of bottle gourd (kg/ha), X1 = Labor (man-days /ha), X2 = tillage (no./ha), X3 = Seed (kg/ha), X4 = bamboo cost (Tk./ha), X5 = cow dung (kg/ha), X6 = Urea (kg/ha), X7 = TSP (kg/ha), X8 = MoP (kg/ha), X9 = Zn (kg/ha), X10 = DAP (kg/ha), X11 = pesticide (liter/ha), X12 = Irrigation (no. of operation/ha), a = Intercept, b1, b2, …………. b12 = Coefficients of the respective variables to be estimated, Ui = Error term. The PCI was computed by using the following formula: PCI= Ph×3 + Pm×2+Pl×1+Pn×0, Where, PCI = Problem Confrontation Index, Ph= Percentage of farmers with “high problem”, Pm= Percentage of farmers with “moderate problem”, Pl= Percentage of farmers with “low problem”, Pn= Percentage of farmers with “no problem”. The PCI of the individual problems could range from 0 to 300, where 0 indicates “no problem” and 300 indicating “high problem” confrontation. Moreover, Pearson’s Correlation Coefficient (r) was used to ascertain the correlation between the diverse characteristics of the bottle gourd farmers and their problem confrontation.