The Yale Food Addiction Scale (YFAS) is the most commonly used scale for measuring food addiction (FA). The previous approach to the YFAS and its subsequent versions assumed dichotomization of items, separating addiction symptoms and clinical significance items, and factorial validity testing on a subset of items. In this paper, we discuss the drawbacks associated with these procedures. In addition, we present a different analytical approach to investigate the validity of the modified YFAS (mYFAS) along with an alternative scoring method that overcomes limitations related to the previous approach. After establishing the structure of the mYFAS, we investigated the potential antecedents and consequences of FA separately for men and women. The sample consisted of 1182 Polish undergraduate students (613 women, 559 men, 10 missing values on gender) with a mean age of 20.33 years (SD = 1.68; range: 18–36). They were asked to complete self-report questionnaires measuring FA, personality traits (Big Five), self-esteem, narcissism, self-efficacy, social anxiety, loneliness, and well-being indicators. Due to the low content, factorial, and clinical validity, the first three items were excluded from the Polish version of the mYFAS. The six-item mYFAS demonstrated measurement invariance, allowing for meaningful comparisons between genders and yielded almost identical prevalence rates for men and women. The hierarchical multiple analysis showed that age, extraversion, narcissism, and social anxiety predicted FA in both genders, whereas important gender differences in antecedents were also noted. In addition, FA was associated with body mass index (BMI) and most of the well-being indicators, even after controlling for relevant variables. The findings suggest that our modified analytical approach allows researchers to measure FA using a valid, useful, and simple tool.