Sixty moderately obese women (mean BMI = 33, mean age = 43), randomized to a lactovegetarian or regular 1300-kcal weight-reducing diet were followed at 3, 8 and 24 months. Weight follow-up was 92%, while 47% complied with the program throughout with no differences between the two diets with respect to compliance rate, weight loss or behavioral test results. Over 24 months compliers lost a mean 3.9 kg compared to a gain of 1.8 kg in the non-compliers. Short-term improvements in mental well-being measured by the Mood Adjective Check List deteriorated after 2 years to lower levels than at entry. Self-assessed motivation to diet was inversely related to mental well-being at two years. Positive long-term changes of functional status (Sickness Impact Profile) were found. Though subjective prediction of success measured after 3 weeks on diet predicted short-term and maximum weight loss, it did not predict ultimate outcome. More difficulties in resisting emotional and social eating cues (high disinhibition score on the Three-Factor Eating Questionnaire) before and during the diet predicted weight gain. The more initial health-related dysfunction (SIP) the greater the weight regain. Psychological characteristics at baseline did not predict compliance or overall weight loss. The magnitude of weight loss after 24 months was related to amount and duration of maximum weight loss.