The objective of this study is to establish a causal relationship between the Mediterranean diet (MD) and various measures of overweightness using the Croatian Adult Health Survey 2003 data. Our results show that among three measures of obesity (body mass index, waist-to-hip ratio (WHR) and obesity (BMI ≥ 30), we found statistically the most convincing relationship between the BMI and the MD. Our results show that an increase in the Mediterranean diet aggregate index by 10% reduces the BMI by about 0.9%. When the MD10 index is replaced with the set of its ten constituent food groups, as a group, these food variables are jointly statistically significant, most of them have expected (negative) signs, and some of them are also individually significant. For the other two overweight measurements (WHR and obesity) we found that the impact of MD aggregate index is insignificant but when the index is replaced by its ten constituent food elements, these are jointly statistically significant in explaining the variation in the obesity measures.