Adjuvant radiotherapy (RT) is routinely recommended for most soft-tissue sarcomas (STS) of the extremities. However, its impact on local control is not clearly established after wide complete excision. We performed a retrospective analysis of patients who underwent wide resection in our institution (first or second resection in cases of incomplete surgery) and either did or did not receive adjuvant RT. All histological specimens of patients operated upon between 1975 and 1996 were carefully analysed and only patients with free tumour margins (ftm) were retained for the analysis. The histopathological classification was as follows: minimal resection (mR) (ftm<10 mm) and optimal resection (oR) (ftm >/=10 mm). There were 133 patients with a median age of 44 years (range 16-88 years). The median tumour size was 6 cm (range 1-20 cm) with 28, 44 and 28% of stage I, II and III lesions, respectively. 93 patients (70%) were reoperated upon and residual tumour was found in 55% of the patients (51/93). 69 patients (17 oR and 52 mR) received adjuvant RT and 64 patients did not (54 oR and 10 mR). Other patient characteristics (age, tumour size, stage, deep-seated lesion, histoprognostic grade, adjuvant chemotherapy) were similar in both the RT and no-RT groups. Median follow-up was 10 years (3-25 years). The 5- and 10-year local relapse-free survival rates were 78 and 71%, respectively. 33 patients relapsed locally: 11 in the RT group and 22 patients in the control group (P=0.01). In the univariate analysis, adjuvant RT was correlated with relapse-free survival, while tumour grade and tumour margin status were correlated with overall survival. The multivariate analysis demonstrated a favourable impact of RT and negative influence of malignant fibrous histiocytoma (MFH) on local relapse-free survival; the tumour grade was correlated with overall survival. RT had a positive influence on local control exclusively in patients with mR resection (P=0.005) and in patients with residual tumour cells after re-excision (P=0.001). RT had no influence on 5- and 10-year overall survival. The 5- and 10-year overall survival for the entire population were 77 and 67%, respectively. Optimal resection seems to be the best predictive parameter for a favourable outcome in localised STS. Adjuvant RT is indicated after mR resection and for residual tumour after definitive surgery, but its role after oR resection (primary resection or no residual tumour after re-excision) should be evaluated in a prospective randomised trial.