Since glioblastomas in adults are uniformly fatal, evaluation of easily reproducible prognostic criteria which would attempt to define groups of patients is required. However, there is lack of a clear consensus regarding the expression of some markers in the literature. Therefore, an immunohistochemical study was performed to determine the prognostic significance of Ki67, p53, and epidermal growth factor receptor (EGFR) in a retrospective series of 63 glioblastomas. Image analysis was carried out in positive specimens to quantify the immunoprecipitates. p53 and EGFR expression were specifically addressed in the 36 primary glioblastomas reported in this series. In all cases, clinical data (age, Karnofsky performance scale index [KPS] before surgery, extent of surgery) and immunohistochemical features were analysed using univariate and multivariate analysis to ascertain whether any significant correlation exists between [1] EGFR expression [2], p53 accumulation [3], Ki67 labelling index and prognosis (survival time and disease-free survival time, DFST). The results showed that in this series of glioblastomas, none of these markers had any prognostic value. Among the clinical parameters, a high KPS before surgery was found to be indicative of a shorter DFST and survival time (P < 0.05), whereas a younger age at onset and total or subtotal surgical excision were associated with a longer survival (P < 0.001 and 0.05, respectively). EGFR protein accumulation was inversely correlated with p53 accumulation (P = 0.01). The percentage of the primary glioblastomas expressing EGFR was much lower in our study (33%) than in the literature suggesting that the molecular distinction between primary and secondary glioblastomas is not so clear-cut.