We report on two randomized trials performed in the USA and Europe, which compared methotrexate and nolatrexed as treatment for patients with recurrent head and neck cancer. Eligibility criteria included: histologically confirmed squamous-cell carcinoma, measurable disease, adequate hematological, renal and hepatic functions, failure of a first-line chemotherapy, and informed consent. Methotrexate 40 mg/m2 was weekly given by short infusion, and nolatrexed 725 mg/m2 per day was administered as a five-day continuous infusion, every three weeks. A total of 139 patients (63 in the USA. 76 in Europe) were randomized based on a ratio of 2/1: 93 and 46 received nolatrexed and methotrexate, respectively. Patient characteristics included 115 males and 24 females; median age 60 years. In the nolatrexed arm, the following grade 3-4 toxicities occurred: neutropenia (29.9%) with 3.1% of febrile neutropenia, mucositis (33.3%), and vomiting (10.3%). In the MTX arm, the grade 3-4 toxicities were neutropenia (7.1%) and mucositis (6.9%). There was no difference in activity between the nolatrexed and the methotrexate treatment: 3.3% and 10.8% of objective responses, 1.9 versus 1.5 months of disease-free progression and 3.5 versus 3.7 months of overall survival, respectively. Nolatrexed has demonstrated a similar activity to methotrexate.