Objective We aimed to study whether dietary fat composition (n-3 and n-6 polyunsaturated fatty acids ratio (PUFAs) and PUFAs and saturated fatty acids (SFAs) ratio) is associated with total body fat (TF) and body fat distribution and whether this association was modified by the presence of chronic disease in middle-aged and elderly women in two population-based cohorts in the Netherlands and Australia. Methods The study was performed in the Rotterdam Study (RS), a prospective cohort study among subjects aged 55 years and older (N = 1182 women) and the Calcium Intake Fracture Outcome Study (CAIFOS), a 5-year randomized controlled trial among women age 70+ (N = 891). At baseline, diet (i.e. PUFAs and SFAs) was measured by validated food frequency questionnaires. TF was assessed using Dual-energy X-ray absorptiometry in both studies and android abdominal fat (AF), gynoid fat (GF) and the android/gynoid ratio (A/G ratio) in the RS but not the CAIFOS. Chronic disease was defined as the presence of cardiovascular disease, diabetes mellitus and cancer. Results No association was found between dietary n-3/n-6 PUFAs ratio or SFA/PUFAs ratio with TF in both cohorts. In the RS, a high n-3/n-6 PUFAs ratio was associated with a higher AF (3rd vs. 2nd tertile (reference): β: 0.15; 95% CI: 0.05, 0.24) but not with the A/G ratio. A low SFA/PUFA ratio was associated with a lower AF (1st vs. 2nd tertile (reference): β: −0.12; 95% CI: −0.22, −0.06) but not with the A/G ratio. Presence of chronic disease was found to be a significant effect modifier in both cohorts with regard to n-3/n-6 PUFAs and TF (P < 0.05). In participants without chronic disease, a higher n-3/n-6 PUFAs ratio was associated with a higher TF in the RS cohort (3rd vs. 2nd tertile (reference): β: 0.94; 95% CI: 0.12, 1.76), but this was not replicated in CAIFOS. Conclusion These findings do not support the hypothesis that dietary fat composition is consistently associated with TF and body fat distribution in women. Future studies should clarify to what extent these findings may be influenced by the presence of chronic disease.