We examined the relationship between postoperative dietary intake (DI) of geriatric hip fracture (HF) patients and their functional and clinical course until 6 months after hospital discharge. In eighty-eight HF patients ≥ 75 years, postoperative DI was estimated with plate diagrams of main meals over four postoperative days. DI was stratified as >50, >25–50, ≤ 25 % of meals served. Functional status according to Barthel index (activities of daily living) and patients' mobility level before fracture, postoperatively, at discharge and 6 months later were assessed and related to DI levels. In-hospital complications were recorded according to clinical diagnosis. Associations were evaluated using χ2 and Kruskal–Wallis tests, and repeated-measures ANOVA and ANCOVA. Postoperatively, 28 % of participants ate >50 %, 43 % ate >25–50 % and 28 % ≤ 25 % of meals served. Irrespective of pre-fracture functional status, patients with DI ≤ 25 % had significantly lower Barthel index scores at all times after surgery (all P< 0·05) and ANOVA revealed a significant time × DI interaction effect (P= 0·047) on development of Barthel index scores that remained significant after adjustment for potential confounders. Patients with DI >50 % more often had regained their pre-fracture mobility level than those with DI ≤ 25 % at discharge (>50 %: 36 %; >25–50 %: 10 %; ≤ 25 %: 0 %; P= 0·001) and 6 months after discharge (88; 87; 68 %; P= 0·087) and had significantly less complications (median 2 (25th–75th percentile 1–3); 3 (25th–75th percentile 2–4); 3 (25th–75th percentile 3–4); P= 0·012). To conclude, geriatric HF patients had very low postoperative voluntary DI and thus need specific nutritional interventions to achieve adequate DI to support functional and clinical recovery.